00:54:21 Monster database of master branch on crawl.develz.org updated to: 0.28-a0-1382-g92835e5c71 00:54:44 Unstable branch on cbro.berotato.org updated to: 0.28-a0-1382-g92835e5c71 (34) 01:48:38 Stable (0.27) branch on crawl.kelbi.org updated to: 0.27.1-34-g3ba077fa99 01:58:13 Stable (0.27) branch on cbro.berotato.org updated to: 0.27.1-34-g3ba077fa99 01:59:59 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-4506-g453a05f5f4 04:22:13 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-4506-g453a05f5f4 10:29:43 Stable (0.27) branch on crawl.akrasiac.org updated to: 0.27.1-34-g3ba077f 10:43:38 03advil02 07[stone_soup-0.26] * 0.26.1-13-g052cdd8: fix: don't let Usk steal other gods' powers (G-Flex) 10(3 weeks ago, 2 files, 31+ 13-) 13https://github.com/crawl/crawl/commit/052cdd8df306 10:59:27 Unstable branch on crawl.akrasiac.org updated to: 0.28-a0-1382-g92835e5 (34) 11:31:59 I'm not sure if this is generally considered something worth solving due to the way ?acq gets "better" the further you are into a game, but would a reasonable way to help reduce its often prolificness in the early game be to prevent it spawning in stacks of 2 or 3 until ≈ D:8? 11:32:41 I also can't remember if ?acq being weirdly common in the early game was something that was discussed in here or in #crawl, but it does seem /weirdly/ common in far too many of my games. 11:32:48 We stopped it spawning in stacks at all, then realised that this incentivises Hypothetically Optimal Man to note stack sizes on all scrolls. I think your proposal has the same downside. 11:33:40 objstat suggested it's not more common (except inasmuch as it didn't used to be able to generate on very early levels at all). I think it seems more common because it is much more likely you got something good, making it more memorable. 11:35:26 hm, true, it's definitely super subject to confirmation bias in that sense 11:36:12 I mean, just in my last game alone I got a D:4 +4 executioner's axe of electrocution which is obviously more memorable... but I also got a whopping /eight/ ?acqs between D:1 and D:12 11:36:38 <10PleasingFungus> my current game had a total of 1 acq through lair 11:36:41 That's definitely on the extreme end of things and certainly quite unlikely, but it still strikes me as odd 11:36:49 <10PleasingFungus> probability is clumpy 🙂 11:38:13 .... my game I'm playing right now (as I'm saying this) I just got a ?acq for a +2 vampiric executioner's axe on D:4. 11:38:21 Probability is /weird/ 11:38:32 <10PleasingFungus> grats 🙂 11:39:52 but if objstat says it's reasonable enough, I suppose I can chalk my weird experience up to confirmation bias. I certainly trust it more than myself :) 11:41:52 I'd run your own objstat to be sure. IDK that some later vanilla change we haven't got hasn't made a bigger difference. 11:48:43 <10PleasingFungus> i'll run a quick 1k iters over d:1-10 11:49:45 (Naturally I died due to me engaging an OOD poorly, but thanks :)) 12:23:21 Stable (0.26) branch on underhound.eu updated to: 0.26.1-13-g052cdd8df3 12:25:12 03advil02 07* 0.28-a0-1383-g7516f22: fix: add a missing Error instance in webtiles 10(76 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/7516f22de055 12:25:12 03advil02 07* 0.28-a0-1384-gb830f16: feat: improve mouse handling for message pane input 10(7 minutes ago, 2 files, 53+ 6-) 13https://github.com/crawl/crawl/commit/b830f16b3f29 12:33:46 Unstable branch on crawl.kelbi.org updated to: 0.28-a0-1384-gb830f16b3f (34) 13:23:46 Stable (0.27) branch on underhound.eu updated to: 0.27.1-34-g3ba077fa99 14:40:08 <13Sastreii> pharaoh ant, steelbarb worm, jorogumo and broodmother. 2 to go 14:40:08 <13Sastreii> https://cdn.discordapp.com/attachments/747522859361894521/926952796144685136/Schermata_2021-01-01_alle_22.34.12.png 14:40:08 <13Sastreii> https://cdn.discordapp.com/attachments/747522859361894521/926952796312453142/lovely_puppies.zip 14:41:47 <08nicolae> q: now that frenzied monsters aren't excluded from stuff like auto-target and cleaving, should they also be targets for auto-attack or whatever tab is called 14:44:54 <05kate> probably, although i have no idea how possible that is in lua 14:48:26 <10PleasingFungus> oh amazing! love the filename and art both 🙂 14:48:32 <10PleasingFungus> will be a little sad to see old steelbarb go 14:52:49 <10PleasingFungus> re earlier acq discussion: looks like acq is now 2-3 times more common in D than it was in 0.26 14:52:49 <08nicolae> speaking of old tiles: should we find someplace to reuse the eye of draining somewhere, since it still appears in the splash screens options iirc 14:52:50 <10PleasingFungus> shrug 14:52:58 <08nicolae> also are jorogumos supposed to be humanoid, what's the deal there 14:53:05 <08nicolae> i mean it looks awesome 14:53:11 ??jorogumo 14:53:11 Jorogumo[1/1]: High-end threat in Spider that casts ensnare & bolt of draining, plus melees for {hornet} poison (i.e. paralysis). From Japanese mythology. New in 0.28. 14:53:16 <08nicolae> aha 14:53:34 hm, I think it was the in-game desc that said they were humanoid 14:53:44 <10PleasingFungus> they're demihumans 14:54:34 <10PleasingFungus> google turns up some reasonable results for them, most of which are safe for work 14:55:27 <10PleasingFungus> @Sastreii i really like the pharaoh ant, broodmother & jorogumo (esp the latter two), but i'm not totally sure about the steelbarb. something about the perspective doesn't quite work for me - it feels very flat 14:55:43 <10PleasingFungus> not sure what other people think 14:56:33 <08nicolae> yeah, tbh, it feels like _ when it should feel more like / imo 14:56:54 <08nicolae> like it's not using as much of the tile space as it could 14:57:25 <10PleasingFungus> it's good for not all tiles to use the entire space 14:57:30 <10PleasingFungus> the more space they take up, the more threatening they feel 14:57:38 <08nicolae> well, true 14:57:42 <08nicolae> it feels... off center, maybe? 14:59:18 <08nicolae> like the balance of the worm is lower in the tile, when smaller stuff is closer to the middle still 14:59:18 <08nicolae> i think 14:59:39 <10PleasingFungus> @kate i'm wondering about turning acq scroll weight back down, since it feels a bit wrong for them to be at 2-3x the older weight - makes em feel less special. WDYT about adding another frequency between 'rare' and 'very rare' and/or renaming 'very rare' to 'extremely rare'? 15:00:34 (phew I'm not crazy) 15:00:37 wait I never played 0.26 15:00:46 darn, not off the hook yet 15:02:54 Unstable branch on crawl.kelbi.org updated to: 0.28-a0-1385-gcb2c1687f3 (34) 15:03:54 <05kate> scrolls don't have any "very rare" entries so they could just use that, but is it definitely related to the weighting standardisation? they went from 14 weight to 15 weight, so it doesn't seem right to me that it'd translate to a 2-3x increase 15:04:17 <05kate> unless something is more significantly broken 15:04:30 03PleasingFungus02 07* 0.28-a0-1385-gcb2c168: Add some lovely puppies (Sastreii) 10(7 minutes ago, 3 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/cb2c1687f3e7 15:05:46 <10PleasingFungus> i haven't bisected it (which would take a while), and i'm not saying it's based on the weighting standardization 15:05:46 <10PleasingFungus> just that, with current weights, acq is a lot more common than it used to be 15:05:46 <05kate> ahh i see yeah 15:05:46 <10PleasingFungus> and it's not based on some new busted vault or something 15:05:46 <05kate> definitely no problem with them going back to being rarer again yeah, although would be interested to know what the cause is in that case 15:06:21 PF: JOOI can I see your objstat results 15:06:24 please? 15:06:31 <10PleasingFungus> sure, gimme a sec 15:07:23 <10PleasingFungus> https://www.dropbox.com/s/2nk5grhg3gjtlxb/objstat_Scrolls.txt?dl=0 should work 15:07:47 <10PleasingFungus> looks like the change might have actually be in 0.27, since the numbers for objstat there look much more similar to my current results than 0.26's 15:12:15 I ask because I ran a 100x objstat today (in our 0.23ish with new acquirement but no changes to scroll drops); 0.42 in D, of which 0.26 on D:11-15. So I think perryprog should have trusted their instincts. 15:13:20 <10PleasingFungus> why not just look at https://drive.google.com/drive/folders/1NV0z4yTXN562U7UOdtxdUlwAvETIJq5z ? 15:14:34 Well, partly I was wondering if new acquirement itself had changed it somehow, which ought to be impossible but... (also, I was running it for the monster stats, I just had it lying around) 15:17:19 03kate-02 07* 0.28-a0-1386-g95c3d9e: Let autofight target frenzied monsters (nicolae) 10(8 minutes ago, 2 files, 12+ 4-) 13https://github.com/crawl/crawl/commit/95c3d9ec3deb 15:18:32 Unstable branch on crawl.kelbi.org updated to: 0.28-a0-1386-g95c3d9ec3d (34) 15:19:05 <13Sastreii> Mmm let’s see what i can do. I’ll try to make another attempt tomorrow prob 15:21:54 <13Sastreii> Ngl pharaoh ant and steelbarb were a pain in the ass to do 15:23:00 <10PleasingFungus> ah, sorry 😦 15:23:28 <13Sastreii> this are some of the attempts i made 😆 15:23:28 <13Sastreii> https://cdn.discordapp.com/attachments/747522859361894521/926963865865052220/Schermata_2021-01-01_alle_23.22.33.png 15:26:49 <10PleasingFungus> wow 15:27:24 <10PleasingFungus> i kinda like bottom-left and bottom-center 15:29:48 I dig the one with the shiny bitey bits 15:32:29 <10PleasingFungus> premium ant 15:36:31 03PleasingFungus02 07* 0.28-a0-1387-g3a22ab9: Restore 0.26 acquirement frequency (perryprog) 10(6 minutes ago, 2 files, 3+ 1-) 13https://github.com/crawl/crawl/commit/3a22ab9bdc81 15:38:00 muahaha what will I get nerfed next 15:38:19 wait I don't want that, I'm bad at this game 15:39:27 !lg perryprog won 15:39:28 No games for perryprog (won). 15:39:38 wait, you play offline, that's right 15:39:47 if you have wins, you're not that bad at this game 15:39:55 I have 4 wins 15:40:15 well, humph. 15:42:01 (I have once died on the orbrun; that's the closest I've ever gotten to a win) 15:42:02 then again, Crawl Is My Solitaire™ 15:47:46 Unstable branch on crawl.kelbi.org updated to: 0.28-a0-1387-g3a22ab9bdc (34) 16:23:01 Unstable branch on underhound.eu updated to: 0.28-a0-1387-g3a22ab9bdc (34) 16:39:00 omg new broodmother tile 16:39:05 like night and day 16:39:23 like crayons and bourbon 17:45:55 <08wormsofcan> isn't acquirement's frequency increase just due to the change hellmonk made allowing them to spawn early 17:46:19 07hellmonk02 {PleasingFungus} * 0.27-a0-663-gc2beb70: Legalize early rare scrolls. 10(10 months ago, 1 file, 12+ 10-) 13https://github.com/crawl/crawl/commit/c2beb70d1432 17:46:19 %git c2beb70d1432f85ddb52b964617ef45e183e7222 17:46:19 That? 17:48:05 <08wormsofcan> yeah 17:52:20 <08wormsofcan> I think there might have also been a change that allowed acquirement to be spawned in stacks when it was limited to 1 at a time previously 17:52:38 <08wormsofcan> if that change exists it would be somewhat old though 17:54:57 07PleasingFungus02 * 0.17-a0-1207-ge31eaae: Allow all scrolls to generate in stacks 10(7 years ago, 9 files, 30+ 42-) 13https://github.com/crawl/crawl/commit/e31eaaee95c4 17:54:57 %git e31eaaee95c42c29122bf2b03cb3ff864740cd7e 17:55:00 Sounds like that 17:55:22 <08wormsofcan> dang that's old 17:55:26 07PleasingFungus02 * 0.17-a0-1206-g6199e19: Remove stack generation restrictions from potions 10(7 years ago, 10 files, 24+ 35-) 13https://github.com/crawl/crawl/commit/6199e197dfb7 17:55:26 %git 6199e197dfb758b78381e7fd94b21fc39cb3421e 17:55:31 That's the previous commit mentioned 17:56:57 I suppose it's a germane point, but personally I read any un-id'd scroll as soon as I'm even mildly safe. There isn't really any actual harm they can do to you in the way !mut and !degen can. The worst that can happen is something like ?noise but those are hardly a threat and you can just read them on a cleared floor if you're paranoid. 17:57:21 So I kinda feel that worrying about spoilerness for scrolls specifically is less important than that of potions 17:58:43 <08wormsofcan> well the harm is reading a blink in a completely safe situation and losing it 17:58:43 <08wormsofcan> if you can deduce which scroll is likely to be blink through spoilers, there is a reason to track spoilers 17:58:43 ah, yeah I guess that's true. I admittedly never really thought of that. 18:52:02 <10PleasingFungus> wormcan: no, it’s a per-floor difference 18:52:10 <10PleasingFungus> not just a difference across all D 18:52:49 <10PleasingFungus> went from 1/20 per-floor chance of acq scroll to something like 1/12 18:53:06 <10PleasingFungus> then the early d stuff added a lil on top 19:20:15 <08wormsofcan> ah, interesting 19:37:08 <06advil> Did anything else change? 19:58:29 <10PleasingFungus> lots yeah 20:00:07 <10PleasingFungus> i didn’t do a full audit 20:00:07 <10PleasingFungus> but all the scrolls i checked were significantly up or down 20:02:42 <08wormsofcan> hmm, this is just 0.26-0.27? 20:04:14 <08wormsofcan> I did a similar analysis of items over a longer period of time because I suspected the removal of several item types would cause an inflation in "good" consumables to show up in stores but that wouldn't explain a shift over the last version 20:15:13 <10PleasingFungus> it was just those two yeah 20:15:33 <10PleasingFungus> and 0.27 seemed close to current values 20:45:34 <08wormsofcan> ok so looking at the numbers a bit more closely, the change could either be shop related or vault related 20:46:30 <08wormsofcan> branches that place shops like orc, vaults and depths have a noticeable bump in acq quantities while lair and zot are the same 20:47:19 <08wormsofcan> but ossuaries also had a very big bump from 0.03 to 0.12, and a lot of ossuary maps have scrolls as loot 20:49:22 <08wormsofcan> swamp is an odd case because there is a bump from 0.2 to 0.26 but it doesn't have any shops, and I don't know much about what kind of loot its vaults place 20:52:18 dang, RARITY_ veryrare, uncommon, and common are all powers of 2 20:52:32 rare is so close to being one as well! 20:53:18 er not powers of 2 20:53:22 squares 20:54:09 :( 20:57:15 New branch created: pull/2328 (8 commits) 13https://github.com/crawl/crawl/pull/2328 20:57:18 03klorpa02 07https://github.com/crawl/crawl/pull/2328 * 0.25-a0-633-g1c9228a: TypoFixes 10(1 year, 9 months ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/1c9228aa1ad1 20:57:18 03klorpa02 07https://github.com/crawl/crawl/pull/2328 * 0.25-a0-634-gc93a2a4: BritishSpellings 10(1 year, 9 months ago, 44 files, 94+ 94-) 13https://github.com/crawl/crawl/commit/c93a2a4c9e6f 20:57:18 03klorpa02 07https://github.com/crawl/crawl/pull/2328 * 0.25-a0-635-g78c0df1: MoreBritishings 10(1 year, 9 months ago, 28 files, 53+ 53-) 13https://github.com/crawl/crawl/commit/78c0df1ef2f5 20:57:18 03klorpa02 {GitHub} 07https://github.com/crawl/crawl/pull/2328 * 0.25-a0-638-gc458211: Merge branch 'master' into master 10(1 year, 9 months ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/c45821171777 20:57:18 03klorpa02 07https://github.com/crawl/crawl/pull/2328 * 0.28-a0-627-gdbdcb76: Merge branch 'master' of https://github.com/crawl/crawl into crawl-master 10(3 months ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/dbdcb76f6d18 20:57:18 03klorpa02 07https://github.com/crawl/crawl/pull/2328 * 0.28-a0-628-g2600a6c: Adding 10(3 months ago, 12 files, 58+ 58-) 13https://github.com/crawl/crawl/commit/2600a6c74bc7 20:57:18 03klorpa02 07https://github.com/crawl/crawl/pull/2328 * 0.28-a0-629-g103c14d: Merge branch 'crawl-master' into master 10(3 months ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/103c14de4c24 20:57:18 03klorpa02 07https://github.com/crawl/crawl/pull/2328 * 0.28-a0-1395-g649d5a9: Merge branch 'master' of https://github.com/crawl/crawl into master 10(10 minutes ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/649d5a977d89 21:03:45 New branch created: pull/2329 (1 commit) 13https://github.com/crawl/crawl/pull/2329 21:03:45 03RojjaCebolla02 {GitHub} 07https://github.com/crawl/crawl/pull/2329 * 0.28-a0-1388-ga7aad4a: Square up scroll rarities 10(6 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/a7aad4ac0464 21:07:05 :) 21:10:12 03klorpa02 07https://github.com/crawl/crawl/pull/2328 * 0.28-a0-1396-ga11fe80: TypoFix 10(3 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/a11fe80c0955 21:14:35 <08wormsofcan> further info on acquirement frequency changes: doesn't seem like it extends to brand weapon scrolls, which are mostly stable from 0.26 to 0.27, but there are some bumps in orc and ossuaries, and weirdly enough in sewers too 21:17:06 <08wormsofcan> lair also has a small increase of about 0.02 on 1-4 21:18:44 I wonder what these would be if vault scroll placement was ignored (or temporary disabled) 21:18:49 compared across versions 21:19:34 Also I think it would be useful to run the same objstat invocation twice to get an idea of the expected variation just from the whim of the RNG 21:20:30 (eg I can well believe that small increase of 0.02 is pure chance) 21:20:45 <08wormsofcan> objstat runs a large number of instances to begin with, it's at least 1000 21:21:23 <08wormsofcan> iirc it's also cpu intensive to run it 21:21:36 Indeed (well, I think the default is (or was?) 100) but I certainly see some variation in 100x runs 21:21:46 <08wormsofcan> no, it's 1000 according to the readme 21:22:02 Perhaps then "was" 21:23:29 <08wormsofcan> also I'm just looking at the swamp/snake sheet at the moment 21:23:29 <08wormsofcan> there are 4 total sheets to account for various playthroughs 21:24:09 <08wormsofcan> it takes 9 hours to generate all 4 datasets according to the readme so I don't know how feasible it is to increase the iteration count 21:24:10 No, the default remains 100 in initfile.cc as of Dec 27 '21 21:24:48 (Perhaps we're talking at cross purposes; there may be some bit of development infrastructure that invokes it with "-iters 1000") 21:27:18 <08wormsofcan> yes, it's in the readme 21:27:32 <08wormsofcan> the files are generated with -iters 1000 21:27:58 <10PleasingFungus> agreed objstat sheets are 1k iters 21:28:00 Right. You're talking about that bit of infrastructure; I was just talking about the behaviour of Crawl 21:28:15 <10PleasingFungus> also agree a difference of 0.02 could be chance 21:28:20 <08wormsofcan> yeah, the numbers I'm reading off are from the objstat sheets in ??objstat 21:31:35 <08wormsofcan> anyway, for acquirement, the consistent outliers in generation other than dungeon is orc and ossuaries 21:32:13 Hm. Orc surprises me, ossuaries makes sense 21:32:39 Though I am interested in floorgod changes since that's where really any earlygame ?acq's come up 21:32:43 <08wormsofcan> orc's bump is specifically in orc:2 21:32:51 <08wormsofcan> about 0.04 21:33:03 Oh, that's quite small 21:33:16 Or at least it sounds small. 21:34:14 <08wormsofcan> well, it's about a 50% increase magnitude wise 21:34:23 <08wormsofcan> previously around 0.8-09, increased to 0.13-0.14 21:34:32 <08wormsofcan> er, 0.08 and 0.09 21:36:21 <08wormsofcan> that one may just be a consequence of item removals 21:36:21 As a comparison, what's something like V:5? 21:36:21 <09gammafunk> there was an objstat question? 21:36:21 so many 21:36:21 <08wormsofcan> V:5's increase is a bit more stable, basically hovering around 0.25 to 0.30 21:36:21 <08wormsofcan> er not increase, just count 21:36:21 <09gammafunk> objstat does use 1k iterations for the "published" data in the learndb topic 21:36:21 <09gammafunk> because it takes a long time to run the 4 jobs 21:36:21 (by the way—what's "??objstat" referring to?) 21:36:32 <09gammafunk> ??objstat 21:36:33 <04Cerebot> objstat[1/2]: Run with "crawl -objstat" in a build of crawl with EXTERNAL_FLAGS_L=-DDEBUG_STATISTICS in your make command (or full debug with "make debug") to generate item/monsters statistics. See crawl -help for the argument details. 21:36:34 <08wormsofcan> yeah there's a question of how much variance there is in the data for objstat 21:36:37 oh, learndb, right 21:37:02 <09gammafunk> oh, well I do also include the sample SD, if that helps, worms 21:37:10 <09gammafunk> for the number field, at least 21:37:11 <08wormsofcan> because I'm trying to see if there are any specific reasons for the increase in acquirement 21:37:12 <09gammafunk> not for all fields 21:38:51 <08wormsofcan> to go back to the acquirement in dungeon frequency question, I think the biggest noticeable change is that on d:5-7, acquirement had a frequency of about 0.2-0.5 while it got bumped up to 0.5-0.8 in 0.27 21:38:54 <10PleasingFungus> the question is more about variance between runs of objstat 21:39:15 <08wormsofcan> I'm not sure what would cause that specific change 21:39:28 <10PleasingFungus> i strongly think you’re overanalyzing 21:39:34 <10PleasingFungus> and essentially p-hacking 21:39:47 <09hellmonk> sewer change is probably due to loot despoiling replacing a bunch of items with dgn.loot_scroll 21:40:13 <09hellmonk> which were formerly like, id/rcurse/some other garbo 21:40:15 <09gammafunk> if you're asking if there's some increase in variance between versions for the objstat data, there shouldn't be, I've used 1k iters for a long time, and the objstat process has been the same 21:40:19 <10PleasingFungus> i do not think this data is consistent or strong enough to draw any conclusions about tiny changes in acq gen between d:5-7 21:40:25 <09gammafunk> in terms of how it collects its data 21:40:27 OTOH if all Lair 1-4 increased by a like amount it seems unlikely _that_'s just chance 21:40:39 <10PleasingFungus> not talking about an increase in variance 21:40:52 <10PleasingFungus> the question is: 21:42:45 <10PleasingFungus> 0.26 data shows 0.02 avg acq generated on some floor. 0.27 shows 0.04 avg there. is that meaningful or noise? 21:42:45 <08wormsofcan> that's not the question, actually 21:42:45 <08wormsofcan> the question is 21:42:45 <08wormsofcan> 0.26 data shows that there is a consistently lower amount of acq generated on early D 21:42:45 <08wormsofcan> but 0.27 shows that the amount is consistent across all D floors 21:42:45 <09gammafunk> worms, isn't that the result of hellmonk's change in allowing certain rare items to generate in early levels? 21:42:45 I think that was a feature before to help startscumming? 21:42:48 ya, that 21:42:51 <10PleasingFungus> no 21:42:54 <08wormsofcan> no, we're talking about floors where acq could already generate 21:43:00 <08wormsofcan> in both versions 21:43:01 ah 21:43:03 <09gammafunk> ok 21:43:03 <10PleasingFungus> d:5 onward 21:43:13 <08wormsofcan> specifically, on d:5-7 acquirement is only around 0.2-0.4 21:43:19 <08wormsofcan> and on the remaining floors it's about 0.5-08 21:43:26 <10PleasingFungus> you’re missing a 0 21:43:33 <08wormsofcan> er yeah 21:43:36 <08wormsofcan> 0.0X instead of 0.X 21:43:50 <09hellmonk> entirely possible that it was still from my change 21:43:52 <08wormsofcan> and in 0.27 acq is consistently around 0.06-0.08 21:44:11 <09hellmonk> there is/was some item code that rolls random2(item level) 21:44:30 <09hellmonk> so changes to that would affect distribution on early-mid d floors 21:44:41 <10PleasingFungus> plausible 21:44:41 <09hellmonk> much more heavily than later 21:44:55 <09gammafunk> well, to be fair, one could look at the means and sds for those floors/versions and just do a simple statistical test, slightly complicated, I guess, by the fact that you're combining data from several floors (unless you wanted to pick one floor). No idea if the difference would be statistically significant 21:45:13 <08wormsofcan> the comparisons are floor by floor 21:45:48 <09gammafunk> sure, can do that, the data are there. 1k floors is still a pretty big sample size 21:46:03 <09gammafunk> but I'm not sure how much it helps even if it is significant, I guess 21:46:12 <09gammafunk> still have to find some underlying cause 21:46:35 <09gammafunk> also wasn't there just a PF commit that reduced acquirement rarity? I assume that's the context for this 21:46:49 <08wormsofcan> also @hellmonk were ossuaries also something that needed to be despoilered? 21:46:51 yes I was complaining that ?acq's are too common 21:47:00 <09hellmonk> yes 21:47:14 <09hellmonk> those probably need some degree of loot nerf too but I haven't looked at them 21:47:18 <08wormsofcan> yeah I was looking into objstat data to try to pinpoint the cause for the increase in acq 21:47:26 <09hellmonk> did a bunch of sewer loot nerfing in the recent revamp 21:47:38 it wouldn't be the worst thing in the world to bisect this, by the way—an -iters 1000 objstat is only around 6 and a half minutes on my laptop 21:47:43 <09hellmonk> whether it's enough or not idk 21:49:10 ossuaries prooobably need loot nerf 21:49:10 they're usually real easy and give lots of consumables 21:49:10 perryprog: you could hack a lot of that by doing the most-affected levels only, surely? 21:49:10 <09gammafunk> you are running it for one floor (or so); this job has to run 4x for the entire dungeon with different dungeon branches in play 21:49:10 Right, I figured as much, but we don't care (as much) about the other branches, right? 21:49:10 Right right right 21:49:10 <10PleasingFungus> imo the problem with ossuaries is more the enemies and less the loot 21:49:10 <09hellmonk> combination of both tbh 21:49:10 <09gammafunk> and yeah you can run objstat with more iters for just your early floors, but you need to compile/run it for 0.26 as well 21:49:22 <08wormsofcan> yeah the other branches are less affected by any changes to acq frequencies 21:49:28 <09gammafunk> mimnay did like 100k (or was it 1 million) iters for D:1 a couple years back 21:49:33 <09gammafunk> which I'm amazed it didn't crash 21:49:42 <08wormsofcan> D, ossuaries and possibly orc:2 are areas with the biggest delta 21:49:46 <09hellmonk> there aren't really a ton of early undead you can place that do things though, like what are you gonna do other than spawn a mummy or sth 21:49:55 <09hellmonk> wights I guess? 21:50:16 <10PleasingFungus> i still want to do ridiculous trap/terrain nonsense 21:50:22 (Also objstat could probably be parallelized but that's a lot of work for such a niche tool) 21:50:23 <08wormsofcan> black mamba zombies :PlogChamp: 21:50:24 <09hellmonk> I think there's still a wraith map but there aren't a lot of characters that can kill a wraith 21:50:48 <09hellmonk> so it's mostly just if you see that one you either kite it around carefully or leave 21:50:57 aww, I've never gotten that one 21:51:04 <10PleasingFungus> fighting extra tough zombies by messing w traps 21:51:11 <09hellmonk> fun fact, it used to be a guardian mummy 21:51:12 there are a few trap-heavy ones 21:51:15 <10PleasingFungus> probably not old traps 21:51:18 <08wormsofcan> zombie_basket 21:51:24 <10PleasingFungus> since those were garbage 21:51:26 but those aren't too interesting since it's just some nets in a chessboard pattern 21:51:28 <09hellmonk> so if you look back far enough you can see people getting obliterated by a guardian mummy in ossuary 21:51:30 <10PleasingFungus> but new and exciting traps 21:51:43 Zot! 21:52:07 <10PleasingFungus> zot is not new or helpful :p 21:52:10 <09hellmonk> the ossuary map where it's just a big hallway and the mummies run at you is ok, since you can't really kite them easily 21:52:23 <09hellmonk> which makes them being slow not so bad 21:52:29 <10PleasingFungus> i want pendulum blades, explosive barrel cacti 21:52:32 <10PleasingFungus> the works 21:52:57 <08wormsofcan> time to remake the indiana jones boulder beetle vault 21:53:04 <09hellmonk> lol 21:53:08 <08wormsofcan> you enter the loot vault and a boulder beetle spawns 21:53:09 <09hellmonk> remaking that vault every version 21:53:34 <09gammafunk> have you not seen my rework of it? 21:53:35 <08wormsofcan> too bad boulder beetle zombies can't roll 21:53:39 <08wormsofcan> yeah I have 21:53:43 <08wormsofcan> I'm just saying, for ossuaries 21:53:43 <09gammafunk> oh ok 21:53:44 They were broken for about 4 years with old boulder beetles 21:53:54 crawl would be vastly improved if it had minigames that suddenly put you in a doom (or whatever that even worse maze thing was) level that you had to win 21:54:07 <08wormsofcan> can you dodge the boulder beetle in the transporter vault though 21:55:32 <08wormsofcan> I don't actually remember the exact layout 21:55:32 <09hellmonk> bring back elephant slugs and make a map where there's an elephant slug zombie behind you and you have to run forward and kill lesser zombies fast enough to not die to the slug 21:55:32 <09gammafunk> it's not a transporter vault 21:55:32 <09gammafunk> there's another bowling alley vault like that though 21:55:32 <10PleasingFungus> hellmonk: yes! yes!! 21:55:32 Why were elephant slugs removed. That's like the most amazing monster name ever. 21:55:32 <08wormsofcan> oh then I might not have seen the vault you're talking about 21:55:32 <10PleasingFungus> @??torpor snail zombie 21:55:32 <04Cerebot> torpor snail zombie (Z) | Spd: 5 | HD: 10 | HP: 73-98 | AC/EV: 6/0 | Dam: 20 | undead, evil, unbreathing | Res: will(13), cold++, poison+++, drown, miasma, neg+++, torm | Vul: holy | XP: 110 | Sz: Large | Int: brainless. 21:55:32 <10PleasingFungus> probably not mean enough…. 21:55:32 <09hellmonk> inspired by the dpeg dungeon entrance with the jelly 21:55:32 <09hellmonk> yeah not quite enough damage on that 21:55:32 <08wormsofcan> I've encountered the boulder beetle transporter vault, don't have any recollection of a boulder beetle vault that isn't a transporter 21:55:32 <10PleasingFungus> perryprog: same reason as agate snails and goliath beetles 21:55:32 <09hellmonk> @??naga warrior zombie 21:55:32 <04Cerebot> naga warrior zombie (Z) | Spd: 8 (move: 140%) | HD: 10 | HP: 71-98 | AC/EV: 4/5 | Dam: 25, 4(constrict) | undead, evil, unbreathing | Res: will(13), cold++, poison+++, drown, miasma, neg+++, torm | Vul: holy | XP: 161 | Sz: Large | Int: brainless. 21:55:34 <09hellmonk> potential ??? 21:55:35 but like... /zombie snail/. That's just so awesome 21:55:37 <10PleasingFungus> @??torpor snail simulacrum 21:55:38 <04Cerebot> torpor snail simulacrum (Z) | Spd: 5 | HD: 10 | HP: 25-36 | AC/EV: 6/0 | Dam: 20(cold:10-29) | undead, evil, unbreathing | Res: will(13), cold+++, poison+++, drown, miasma, neg+++, torm | Vul: fire, holy | XP: 51 | Sz: Large | Int: brainless. 21:55:44 <10PleasingFungus> too fragile 21:55:47 <09hellmonk> simu might be too frail yeah 21:55:50 <10PleasingFungus> i like your naga 21:55:51 <09hellmonk> can just blow it up 21:55:54 PF: FWIW I just did two 1000x objstat runs for Lair; two levels vary by 0.01, the others don't. So I think a change of 0.02 on all of Lair:1-4 is unlikely to be chance. 21:56:07 <10PleasingFungus> thanks! 21:56:13 <10PleasingFungus> that’s helpful 21:56:14 <09hellmonk> someone will try hexing it and die to the wakeup 21:56:22 <09hellmonk> they will get "mixed" 21:56:26 <10PleasingFungus> lol 21:56:32 <09gammafunk> @wormsofcan 21:56:33 <09gammafunk> https://cdn.discordapp.com/attachments/747522859361894521/927062786998149120/unknown.png 21:56:41 <09gammafunk> that's the indiana jones one 21:56:54 <09gammafunk> referencing the first movie, and replacing the old one 21:57:00 <09hellmonk> also you gotta make sure the bad zombie shows up as extremely dangerous 21:57:06 <09gammafunk> note the hat and whip :gammafHeh: 21:57:07 <10PleasingFungus> yes 21:57:08 <09hellmonk> so people dont just autopilot tab and instantly die 21:57:10 <08wormsofcan> I might have seen that one, but I have never entered it 21:57:15 <09gammafunk> shame 21:57:19 <10PleasingFungus> @??nagaraja zombie 21:57:20 <04Cerebot> nagaraja zombie (Z) | Spd: 8 (move: 140%) | HD: 15 | HP: 108-155 | AC/EV: 4/5 | Dam: 21, 5(constrict) | undead, evil, unbreathing | Res: will(20), cold++, poison+++, drown, miasma, neg+++, torm | Vul: holy | XP: 462 | Sz: Large | Int: brainless. 21:57:29 <10PleasingFungus> that’d do it 21:57:29 <09gammafunk> you could have had an artefact jewel 21:57:36 <08wormsofcan> we need a stetson variant for hat tiles so the vault is represented better in tiles 21:57:41 <09gammafunk> what about naga warrior zombie? 21:57:48 <09hellmonk> huge xp if they can kill it which implies the map needs to be fairly short 21:57:50 <08wormsofcan> or whatever the brand of hat indiana wears is called 21:57:54 <09gammafunk> makes me sad to make zombies of big spell casters 21:57:55 <10PleasingFungus> gamma: that was hellmonk’s suggestion 21:58:00 <10PleasingFungus> possibly more reasonable 21:58:08 <09gammafunk> oh, sorry 21:58:09 <10PleasingFungus> i’m just playing w variant options 21:58:10 also, possible sprint? 21:58:10 <09hellmonk> probably just a couple hallways 21:58:10 <08wormsofcan> oh it's an actual fedora 21:58:18 if you want to take it the whole way 21:58:31 <09gammafunk> fedora tile would be great, yeah 21:58:33 <09hellmonk> this is a very gimmicky map though, so somebody else can make it 21:58:37 <10PleasingFungus> haha 21:58:50 <09gammafunk> pf likes making vaults! he told me as much! 21:58:55 <10PleasingFungus> perry: there is immense untapped potential for crawl sprints 21:58:59 <08wormsofcan> well, hats could use a revamp anyway 21:59:01 that's for sure 21:59:03 <08wormsofcan> I guess I'll work on that 21:59:12 <09gammafunk> cool 21:59:20 <09hellmonk> when was the most recent sprint added, like 0.16 21:59:21 <10PleasingFungus> @gammafunk i dramatically increased my vault count last month! 21:59:27 <09gammafunk> ah, yes you did 21:59:46 <10PleasingFungus> !lg * kmap~~pf_ week 21:59:47 <04Cerebot> 21. 1c5c the Acrobat (L25 DjCj of Gozag), mangled by a spectral caustic shrike (kmap: pf_hell_entry_tar_river) on Depths:5 on 2022-01-01 09:29:16, with 439866 points after 70824 turns and 5:36:30. 21:59:48 <09hellmonk> I thought about making one but then realized it would take like 5x as long as an encompass vault for the main game and less people would play it 21:59:55 <10PleasingFungus> ehh 22:00:02 <10PleasingFungus> i think they’re quite popular when added 22:00:10 <10PleasingFungus> and occasionally thereafter 22:00:25 <10PleasingFungus> !lg * kmap~~pf_ week s=kmap 22:00:26 <04Cerebot> 21 games for * (kmap~~pf_ week): 7x pf_orc_diabolical, 3x pf_guarded_unrand_delatra, 3x pf_guarded_unrand_woodcutter, 2x pf_hell_entry_tar_river, 2x pf_snail_temple, pf_eden, pf_guarded_unrand_meek, pf_torpor_hydra, pf_volcano_lost_world 22:01:50 <10PleasingFungus> good ol diabolical 22:01:50 <09hellmonk> need to finish my spider ends smh 22:01:50 <09hellmonk> maybe tomorrow 22:01:50 <10PleasingFungus> unrands already getting kills, unsurprisingly 22:01:50 Meatsprint is amazing 22:01:50 <10PleasingFungus> it is 22:01:50 I proudly have a win (or two?) with the secret bonus rune (using a cheese strat, of course) 22:01:50 <09gammafunk> orc diabolical, hard to top that one 22:01:50 <09hellmonk> hmm, you should make the opposite of meatsprint 22:01:50 <09gammafunk> it even killed mikee.... 22:01:50 <09hellmonk> vegetablesprint 22:01:50 <08wormsofcan> vegansprint 22:01:50 <09hellmonk> you have to go ely and pacify every monster 22:01:50 <10PleasingFungus> wow 22:02:03