00:01:32 <gammafunk> !lg !bot !@goodplayerspre19 t0.19 !boring / won
00:01:44 <gammafunk> made that nick with !nick goodplayerspre19 $(!lg * name=$(!lg * t0.19 s=name fmt:"${.}" join:"|" title:"") start<'2016-11-04 20:00:00' title:"" s=name fmt:"${.}" join:" " / won ?: N>9)
00:02:12 <Sequell> 441/62953 games for bot (!@goodplayerspre19 t0.19 !boring): N=441/62953 (0.70%)
00:02:27 <gammafunk> it's not really what is says on the tin, since it's only players that participated in 0.19 tourney, but that's what we're looking at now
00:02:30 <gammafunk> !lg !bot !@goodplayers t !boring / won
00:02:52 <gammafunk> goodplayers has been updated since the tourney started, so it's also not 100% accurate
00:03:22 <gammafunk> to the extent that looking at non-goodplayers is an accurate way to gage this in the first place
00:03:30 <Sequell> 305/39859 games for bot (!@goodplayers t !boring): N=305/39859 (0.77%)
00:05:18 <elliptic> !lg !bot !@goodplayerspre19 t0.19 end<'2016-11-15 00:00:00' !boring / won
00:05:28 <gammafunk> !lg !bot !@goodplayerspre20 t !boring / won
00:05:33 <gammafunk> made that nick the same way just now
00:05:40 <gammafunk> but with the 0.20 t date
00:05:51 <Sequell> 269/41501 games for bot (!@goodplayerspre19 t0.19 end<'2016-11-15 00:00:00' !boring): N=269/41501 (0.65%)
00:06:09 <Sequell> 371/40580 games for bot (!@goodplayerspre20 t !boring): N=371/40580 (0.91%)
00:06:12 <elliptic> gammafunk: 0.65 vs 0.77 is still pretty large
00:06:14 <gammafunk> yeesh
00:06:18 <gammafunk> it's even wore, yeah
00:06:27 <gammafunk> if you look at goodplayers defined in the same way for 0.20
00:06:33 <gammafunk> s/wore/worse/
00:06:43 <elliptic> or 0.65 vs 0.91, yeah
00:07:13 <gammafunk> almost 50% higher for non-goodplayers
00:07:49 <elliptic> 40%, yes
00:07:51 <advil> it doesn't appear evenly distributed across species (though who knows what would past significance testing). Biggest boost for Ha, Mi, Ce, HO, Mf, Tr, Vs, Dg, Hu.
00:08:05 <advil> so that makes it look melee related somehow
00:08:33 <advil> s/past/pass
00:08:40 <elliptic> probably we don't have the sample size for individual species to be very significant but I agree that looks vaguely meleeish
00:08:54 <elliptic> I also noticed this:
00:08:58 <elliptic> !lg * t !boring !@bot xl>=9 / xl>=18
00:09:01 <Sequell> 2635/14369 games for * (t !boring !@bot xl>=9): N=2635/14369 (18.34%)
00:09:22 <elliptic> !lg * t0.19 end<'2016-11-15 00:00:00' !boring !@bot xl>=9 / xl>=18
00:09:32 <Sequell> 2252/14319 games for * (t0.19 end<'2016-11-15 00:00:00' !boring !@bot xl>=9): N=2252/14319 (15.73%)
00:10:04 <elliptic> but fairly close to equal for before XL 9 and after XL 18
00:10:18 <cut1less> funny guess: draconian ghosts can't get draining breath anymore. especially online, that could make a big difference, right?
00:10:55 <elliptic> I wouldn't expect that to be anywhere near this large
00:11:09 <gammafunk> average no. haste pots up about 1.85 on average,  no hw pots up about 2.2
00:11:10 <elliptic> !lg * t ghostly dr
00:11:12 <Sequell> 1704. kickascii the Fighter (L15 DrTm of Hepliaklqana), blasted by an ogre mage (crystal spear) on Orc:2 (mines4_lemuel) on 2017-06-06 03:54:22, with 86256 points after 27676 turns and 1:34:53.
00:11:17 <elliptic> gammafunk: for the entire game?
00:11:21 <gammafunk> yes
00:11:44 <gammafunk> alllevels for 3-rune branches (swamp-snake), but including all non-zig portals as well
00:12:05 <gammafunk> so winning games won't see quite that large an increase
00:12:11 <elliptic> hard to imagine that helping more than good wand removal hurt
00:12:27 <advil> well I guess the question is when they show up, but that's not as big a difference as I expected
00:13:06 <gammafunk> yeah, looking at drops in D specifically, the difference is less than one pot for each
00:13:10 <gammafunk> much less than one pot for haste
00:14:30 <elliptic> and it's not like good wands only showed up in late game previously
00:14:59 <gammafunk> I almost suspect it might be the wand changes maybe, but that's pretty hard to prove
00:15:36 <gammafunk> clouds/scattershot are really strong, lightning rod evocable might be somewhat more common than the rod was
00:16:12 <elliptic> I wasn't that impressed by scattershot but I didn't test it a huge amount
00:23:07 <advil> !lg * t0.19|t cv=0.19|0.20 s=cv x=avg(sdam),median(sdam),std(sdam)
00:23:15 <Sequell> 138547 games for * (((t0.19 || t)) cv=0.19|0.20): 85238x 0.19 [10.24, 7, 10.65], 53309x 0.20 [10.83, 8, 11.06]
00:24:21 <gammafunk> what about achievable AC/EV/SH?
00:24:28 <gammafunk> are those easier to get to a higher level this ver
00:25:07 <gammafunk> !lg * t0.19|t cv=0.19|0.20 s=cv x=avg(ac),median(ac),std(ac)
00:25:16 <Sequell> 138553 games for * (((t0.19 || t)) cv=0.19|0.20): 85238x 0.19 [8.08, 5, 8.81], 53315x 0.20 [8.4, 5, 9.38]
00:25:21 <gammafunk> also I guess mhp
00:25:35 <gammafunk> !lg * t0.19|t cv=0.19|0.20 s=cv x=avg(mhp),median(mhp),std(mhp)
00:25:43 <Sequell> 138556 games for * (((t0.19 || t)) cv=0.19|0.20): 85238x 0.19 [51.96, 37, 47.04], 53318x 0.20 [54.42, 38, 50.52]
00:25:57 <advil> !lg * t0.19|t cv=0.19|0.20 won s=cv x=avg(ac),median(ac),std(ac)
00:26:17 <gammafunk> yeah was going to say, maybe looking at won games is better
00:26:27 <advil> for sdam that doesn't work :)
00:26:42 <gammafunk> yes, I mean in response to my query
00:27:23 <elliptic> probably best not to restrict to cv=0.19|0.20 instead of 0.19-a/0.20-a
00:27:32 <elliptic> since that just means something about different servers
00:27:41 <gammafunk> yes, that's true, especially in 0.19
00:27:46 <advil> I was assuming that -a means the server is misconfigured?
00:27:49 <gammafunk> we didn't have cwz on 0.19.0 for some time
00:27:54 <advil> or was for a little while
00:27:56 <advil> ah
00:28:37 <gammafunk> obv we need a "game balance" make test that we plug into travis...
00:28:46 <gammafunk> can't imagine that would lead to false alarms!
00:29:17 <gammafunk> <travis> The build is nerfed
00:31:06 <elliptic> gammafunk: there are actually more 0.20-a games this t than last one
00:31:20 <elliptic> like 2000 games on CAO
00:31:35 <gammafunk> hrm
00:31:43 <elliptic> 0.20.0-b1 counts has 0.20-a
00:31:50 <elliptic> so it doesn't mean that much
00:32:11 <gammafunk> yeah, maybe my recollection of how CAO got added is a bit hazy, but that seems like that shouldn't be the case
00:32:47 <gammafunk> cao got 0.20.0 before the tournament started
00:33:03 <gammafunk> or maybe it didn't and I'm just misremembering
00:33:11 <gammafunk> !lg * cao t cv=0.20-a
00:33:12 <Sequell> 2084. GnoME the Conqueror (L27 GrFi of The Shining One), escaped with the Orb and 15 runes on 2017-05-28 11:38:47, with 12518579 points after 130781 turns and 14:27:27.
00:33:24 <gammafunk> !lg * cao t cv=0.20-a x=vlong
00:33:25 <Sequell> 2084. [vlong=0.20-b1-2-gc5a9893] GnoME the Conqueror (L27 GrFi of The Shining One), escaped with the Orb and 15 runes on 2017-05-28 11:38:47, with 12518579 points after 130781 turns and 14:27:27.
00:33:55 <johnny0> !tstats 10 t tiles=false
00:34:01 <Sequell> Stats after 10 days (t tiles=false): 215 players, 68 runers, 41 winners, 124 wins, 2910 games, winrate 4.26%, total player time 95d+2:38:19.
00:34:14 <johnny0> !tstats 10 t0.19 tiles=false
00:34:28 <gammafunk> !tstats 10 t0.19 tiles=false
00:34:28 <gammafunk> doh
00:34:33 <johnny0> rip sequell
00:35:12 <gammafunk> fairly small number of players there
00:35:12 <Sequell> Stats after 10 days (t0.19 tiles=false): 237 players, 85 runers, 43 winners, 107 wins, 3995 games, winrate 2.68%, total player time 100d+10:58:53.
00:35:56 <elliptic> it's sort of weird, whenever I do queries about XL 9-17 for t and t0.19 the most noticeable thing is that two-headed ogres are very dangerous now
00:36:04 <elliptic> but that would point towards things being harder
00:36:12 <Sequell> Stats after 10 days (t0.19 tiles=false): 237 players, 85 runers, 43 winners, 107 wins, 3995 games, winrate 2.68%, total player time 100d+10:58:53.
00:36:30 <gammafunk> would kind of point or changes in Lair/Orc
00:36:46 <gammafunk> changes that would affect survival in those areas
00:37:11 <gammafunk> I guess 9-12 is still solidly dungeon
00:37:41 <advil> two-headed ogres replaced hill giants
00:37:51 <Doesnty> hill giants disappeared entirely, didn't they?
00:37:52 <elliptic> advil: yeah but they are much more dangerous
00:37:56 <Doesnty> from the spawn table
00:38:07 <n1> well they don't get throwing nets
00:38:08 <advil> ah...yeah they do feel more dangerous
00:38:15 <n1> but of course they're stronger otherwise
00:38:21 <Doesnty> also wands of scattershot exist now
00:38:21 <gammafunk> true, the throwing net change
00:38:58 <elliptic> !lg * t0.19 ckiller=hill_giant s=place
00:39:07 <Sequell> 485 games for * (t0.19 ckiller=hill_giant): 177x D:10, 63x D:11, 45x D:9, 43x D:7, 41x D:6, 35x D:8, 26x D:5, 24x D:12, 13x D:13, 4x D:15, 4x Orc:1, 3x D:14, 2x D:4, Vaults:4, Temple, Shoals:1, Orc:2, D:1
00:39:10 <gammafunk> the lightning rod evocable is a lot more common than old lightning rod was
00:39:13 <elliptic> !lg * t0.19 ckiller=two-headed_ogre s=place
00:39:21 <Sequell> 135 games for * (t0.19 ckiller=two-headed_ogre): 29x D:8, 22x D:9, 12x D:5, 12x Orc:2, 10x D:10, 9x D:3, 7x D:4, 6x D:6, 6x D:7, 5x Vaults:1, 5x Orc:1, 4x D:11, 2x D:12, Vaults:4, D:13, D:14, Volcano, Depths:1, Vaults:2
00:39:22 <elliptic> !lg * t ckiller=two-headed_ogre s=place
00:39:23 <Sequell> 1298 games for * (t ckiller=two-headed_ogre): 505x D:8, 201x D:9, 105x D:7, 90x D:10, 85x D:6, 79x D:5, 64x D:4, 40x D:3, 30x D:11, 29x Orc:2, 15x Orc:1, 14x D:12, 11x D:15, 6x Volcano, 5x Vaults:2, 4x D:13, 4x Vaults:1, 3x D:14, 2x Vaults:3, 2x D:2, 2x Vaults:4, Depths:2, Temple
00:39:24 <gammafunk> but misc drops overall are not up much compared to two versions
00:39:43 <elliptic> over twice as many kills by two-headed ogre now as both monsters combined last version
00:39:50 <elliptic> in 11 days vs 16, too
00:40:49 <elliptic> the hill giant kills were a little later, which makes them a bit more important, but still that looks like the change probably made the game harder
00:43:29 <gammafunk> !lm !bot t br.enter=lair / won
00:43:31 <Sequell> 1156/7832 milestones for bot (t br.enter=lair): N=1156/7832 (14.76%)
00:43:34 <gammafunk> !lm !bot t0.19 br.enter=lair start<'2016-11-15 00:00:00' / won
00:43:40 <Sequell> 1082/7852 milestones for bot (t0.19 br.enter=lair start<'2016-11-15 00:00:00'): N=1082/7852 (13.78%)
00:43:51 <gammafunk> looks like there's not a huge difference in the post-lair-entry winrate
00:44:11 <elliptic> !lm !boring !@bot t br.enter=lair !alive / won
00:44:13 <Sequell> 1156/7428 milestones for boring (!@bot t br.enter=lair !alive): N=1156/7428 (15.56%)
00:44:18 <elliptic> need to remove the games in progress
00:44:21 <alexjurkiewicz> it owuld be interesting to add to travis some tests of things like: fsim of certain hardcoded encounters
00:44:37 <gammafunk> oh, true
00:44:38 <elliptic> !lm !bot !boring t0.19 br.enter=lair start<'2016-11-15 00:00:00' / won
00:44:50 <Sequell> 1082/7736 milestones for bot (!boring t0.19 br.enter=lair start<'2016-11-15 00:00:00'): N=1082/7736 (13.99%)
00:45:11 <elliptic> 15.56 vs 13.99 is a decent amount of the discrepancy
00:47:12 <gammafunk> what's the best way to define that descrepancy, here? I was thinking more in terms of winrate overall being up but 25% (2% -> 2.5% or so)
00:47:49 <gammafunk> but I'm not sure how to reason about that when we start dividing the winrates into XL categories
00:48:08 <gammafunk> s/but 25%/by 25%/
00:50:02 <elliptic> hm I guess I was misremembering how large the winrate change was
00:50:24 <elliptic> it was 1.79% -> 2.27%
00:50:38 <elliptic> !calc 2.27/1.79
00:50:39 <Sequell> 1.27
00:50:52 <elliptic> so yeah, about 25%
00:51:32 <elliptic> !calc 15.56/13.99
00:51:33 <Sequell> 1.11
00:52:00 <elliptic> still pretty close to half the discrepancy
00:52:04 <gammafunk> yep
00:54:06 <advil> !lg * cv=0.19-a|0.20-a s=cv / won
00:54:08 <Sequell> 9206/1073158 games for * (cv=0.19-a|0.20-a): 4797/507014x 0.19-a [0.95%], 4409/566144x 0.20-a [0.78%]
00:54:57 <elliptic> !lg * cv=0.19-a|0.20-a !boring !@bot s=cv / won
00:55:13 <Sequell> 9070/923611 games for * (cv=0.19-a|0.20-a !boring !@bot): 4671/432149x 0.19-a [1.08%], 4399/491462x 0.20-a [0.90%]
00:55:23 <advil> time to bisect
00:55:57 <elliptic> !lg * cv=0.19-a|0.20-a !boring !@bot s=cv / urune>0
00:56:05 <Sequell> 31688/923611 games for * (cv=0.19-a|0.20-a !boring !@bot): 16231/491462x 0.20-a [3.30%], 15457/432149x 0.19-a [3.58%]
00:56:26 <elliptic> !lg * cv=0.19-a|0.20-a !boring !@bot urune>0 s=cv / won
00:56:29 <Sequell> 9070/31688 games for * (cv=0.19-a|0.20-a !boring !@bot urune>0): 4671/15457x 0.19-a [30.22%], 4399/16231x 0.20-a [27.10%]
00:57:11 <elliptic> that sort of drop in winrate is what I would have expected looking at the 0.20 changelog
00:57:25 <advil> yeah
00:57:32 <gammafunk> hrm
00:57:36 <elliptic> weird that the tourney is the other direction, maybe that really does point to it being caused by a change in tourney playerbase composition
00:57:44 <elliptic> or how much time players have in tourney
00:57:55 <elliptic> it's still a really large effect for stuff like that though
00:58:02 <gammafunk> yeah, although that goodplayers query didn't seem to support that
00:58:12 <gammafunk> in that winrates were up in 0.20 for non-goodplayers
00:58:48 <gammafunk> !lg !bot !@goodplayerspre20 !boring t / won
00:59:06 <Sequell> 371/40737 games for bot (!@goodplayerspre20 !boring t): N=371/40737 (0.91%)
00:59:16 <gammafunk> !lg !bot !@goodplayerspre19 !boring t0.19 start<'2016-11-15 00:00:00' / won
00:59:36 <Sequell> 321/41994 games for bot (!@goodplayerspre19 !boring t0.19 start<'2016-11-15 00:00:00'): N=321/41994 (0.76%)
00:59:53 <gammafunk> !calc 91.0 / 6
00:59:54 <Sequell> 15.17
00:59:55 <gammafunk> !calc 91.0 / 96
00:59:56 <Sequell> 0.95
00:59:58 <gammafunk> dammit
01:00:01 <gammafunk> !calc 91.0 / 76
01:00:02 <Sequell> 1.2
01:00:10 <elliptic> alternatively the difference between the general 0.19-a vs 0.20-a stats and the tourney stats could mean that there is a fairly recent bug
01:00:40 <elliptic> !lg !bot !@goodplayerspre19 !boring t0.19 end<'2016-11-15 00:00:00' / won
01:01:13 <Sequell> 269/41501 games for bot (!@goodplayerspre19 !boring t0.19 end<'2016-11-15 00:00:00'): N=269/41501 (0.65%)
01:01:17 <elliptic> gammafunk: most queries should use end rather than start for the time cutoff
01:01:36 <elliptic> since for !lg queries for current tourney we are just getting games that have ended by now
01:01:46 <advil> !lg * 0.20-a|0.20 s=week(end) / won -graph
01:01:48 <elliptic> !calc 91.0/65
01:01:48 <Sequell> 5559/621623 games for * (0.20-a || 0.20): https://shalott.org/graphs/44109ca1751d2d53f33fdfacef214b227e2640f7.html
01:01:49 <Sequell> 1.4
01:01:56 <gammafunk> I see what you mean
01:02:42 <gammafunk> er
01:02:46 <gammafunk> may 21?
01:03:07 <gammafunk> I'm no scientist, but...
01:03:24 <advil> well winrates shoot up for t
01:03:47 <gammafunk> t started may 26
01:04:03 <advil> I think the way sequell does it is that is the week beginning may 21
01:04:08 <gammafunk> oh ok
01:04:11 <advil> so including the first few days of t
01:04:19 <gammafunk> maybe in this case use start?
01:04:21 <elliptic> winrate always does that during t
01:04:45 <elliptic> !lg * ((0.19-a || 0.19)) s=week(end) / won -graph
01:04:48 <Sequell> 7754/872328 games for * (0.19-a || 0.19): https://shalott.org/graphs/1b7546a2d72658fc407d4a13d9e431d477d059df.html
01:04:55 <advil> start vs end is the start or end of the game. not sure if there's a way to get sequell to label the weeks differently
01:05:21 <gammafunk> yeah it's the same bump
01:05:35 <gammafunk> 5 days before
01:05:56 <advil> could do it by day too but I wasn't sure there'd be enough games in -a for it to be meaningfull
01:06:05 <advil> *meaningful
01:06:10 <elliptic> !lg * 0.20-a|0.20 !boring !@bot s=week(end) / won -graph
01:06:17 <Sequell> 5531/542116 games for * (((0.20-a || 0.20)) !boring !@bot): https://shalott.org/graphs/eb8d1f6d014aaab91d5cd5d015df0846b3f01a95.html
01:06:37 <elliptic> !lg * 0.19-a|0.19 !boring !@bot s=week(end) / won -graph
01:06:44 <Sequell> 7615/767850 games for * (((0.19-a || 0.19)) !boring !@bot): https://shalott.org/graphs/31cb473bfdb38421bb6cdca3cfbbb348f82c225b.html
01:07:40 <advil> !lg * 0.20-a|0.20 !boring !@bot s=day(end) / won -graph:area
01:07:47 <Sequell> 5531/542117 games for * (((0.20-a || 0.20)) !boring !@bot): https://shalott.org/graphs/bec67d7fa2689af3daebf72ff1d79c4548ec4645.html
01:08:12 <elliptic> the jump from april 16 to april 23 is slightly suspicious in the week view I guess
01:09:05 <advil> !lg * 0.19-a|0.19 !boring !@bot s=day(end) / won -graph:area
01:09:12 <Sequell> 7615/767850 games for * (((0.19-a || 0.19)) !boring !@bot): https://shalott.org/graphs/1f1fceea9259cfccb4c9d4be4067fe698d5d2ab8.html
01:09:58 <gammafunk> I see spell fail rate breakpoint changes in that range, which was a big set of commits
01:10:50 <advil> yeah, I was worried about that, but I'd expect a different pattern of what chars it affects
01:10:50 <gammafunk> the other commits don't seem like they'd have much in the way of gameplay effects
01:11:08 <gammafunk> !tstats 10 mage
01:11:14 <gammafunk> er
01:11:15 <Sequell> Stats after 10 days (mage): 12 players, 0 runers, 0 winners, 0 wins, 64 games, winrate 0.00%, total player time 18:29:37.
01:11:19 <gammafunk> !tstats 10 t mage
01:11:30 <gammafunk> !tstats 10 t0.19 mage
01:11:32 <Sequell> Stats after 10 days (t mage): 1352 players, 258 runers, 132 winners, 245 wins, 12092 games, winrate 2.03%, total player time 298d+7:18:35.
01:12:12 <gammafunk> I suppose that should include warriormage as well
01:12:25 <gammafunk> !tstats 10 t mage|warriormage
01:12:30 <gammafunk> !tstats 10 t0.19 mage|warriormage
01:12:31 <Sequell> Stats after 10 days (t0.19 mage): 1361 players, 231 runers, 116 winners, 211 wins, 11700 games, winrate 1.80%, total player time 260d+19:14:32.
01:12:47 <Sequell> Stats after 10 days (t mage|warriormage): 1632 players, 340 runers, 177 winners, 403 wins, 19513 games, winrate 2.07%, total player time 1y+77d+17:50:29.
01:13:48 <Sequell> Stats after 10 days (t0.19 mage|warriormage): 1662 players, 308 runers, 152 winners, 334 wins, 20072 games, winrate 1.66%, total player time 1y+24d+2:16:04.
01:14:32 <gammafunk> !tstats 10 t warriormage
01:14:36 <gammafunk> !tstats 10 t0.19 warriormage
01:14:44 <Sequell> Stats after 10 days (t warriormage): 980 players, 178 runers, 100 winners, 158 wins, 7424 games, winrate 2.13%, total player time 144d+10:41:14.
01:14:50 <gammafunk> !tstats 10 t warrior|adventurer|zealot
01:14:59 <gammafunk> !tstats 10 t0.19 warrior|adventurer|zealot
01:15:43 <Sequell> Stats after 10 days (t0.19 warriormage): 1001 players, 161 runers, 87 winners, 123 wins, 8372 games, winrate 1.47%, total player time 128d+7:01:32.
01:16:46 <Sequell> Stats after 10 days (t0.19 warrior|adventurer|zealot): 2191 players, 737 runers, 354 winners, 607 wins, 31571 games, winrate 1.92%, total player time 1y+272d+1:05:18.
01:17:02 <Sequell> Stats after 10 days (t warrior|adventurer|zealot): 2057 players, 779 runers, 383 winners, 721 wins, 29730 games, winrate 2.43%, total player time 1y+281d+5:44:57.
01:19:03 <advil> hmm
01:21:45 <gammafunk> doesn't really seem connected to class too much, or at least not to spells, since it's more increased for non-mages
01:21:56 <advil> yeah
01:22:15 <advil> I was also just checking in PM if there's any class correlation with D:1 deaths and there doesn't seem to be
01:23:03 <advil> !lg * t0.19|t !@bot !boring s=cv mage / place=d:1
01:23:10 <advil> !lg * t0.19|t !@bot !boring s=cv warrior / place=d:1
01:23:15 <Sequell> 7967/29982 games for * (((t0.19 || t)) !@bot !boring mage): 4744/17288x 0.19 [27.44%], 3076/12072x 0.20 [25.48%], 129/538x 0.20-a [23.98%], 18/84x 0.19-a [21.43%]
01:23:27 <Sequell> 4042/50122 games for * (((t0.19 || t)) !@bot !boring warrior): 2518/30803x 0.19 [8.17%], 1443/18171x 0.20 [7.94%], 65/931x 0.20-a [6.98%], 16/217x 0.19-a [7.37%]
01:23:54 <advil> well, the difference is slightly bigger for mages I guess, I didn't do it with !boring before
01:24:30 <advil> !lg * t0.19|t !@bot !boring s=cv warriormage / place=d:1
01:24:48 <Sequell> 5799/20898 games for * (((t0.19 || t)) !@bot !boring warriormage): 3619/12922x 0.19 [28.01%], 2041/7390x 0.20 [27.62%], 77/310x 0.20-a [24.84%], 62/276x 0.19-a [22.46%]
01:27:52 <advil> !lg * vlong<0.20-a0-1022-g99c12e3 !t / won
01:27:56 <Sequell> 55374/7234571 games for * (vlong<0.20-a0-1022-g99c12e3 !t): N=55374/7234571 (0.77%)
01:27:59 <advil> !lg * vlong>0.20-a0-1022-g99c12e3 !t / won
01:28:01 <Sequell> 1014/125768 games for * (vlong>0.20-a0-1022-g99c12e3 !t): N=1014/125768 (0.81%)
01:28:03 <advil> oh heh
01:28:10 <advil> !lg * 0.20-a vlong<0.20-a0-1022-g99c12e3 !t / won
01:28:12 <Sequell> 3596/466848 games for * (0.20-a vlong<0.20-a0-1022-g99c12e3 !t): N=3596/466848 (0.77%)
01:28:29 <advil> (that's the vlong for the main fail rate change)
01:43:30 <gammafunk> !lg !bot !@goodplayerspre20 !t cv=0.20-a / won
01:43:40 <gammafunk> !lg !bot !@goodplayerspre20 t cv>=0.20-a / won
01:44:09 <Sequell> 2460/509965 games for bot (!@goodplayerspre20 !t cv=0.20-a): N=2460/509965 (0.48%)
01:44:28 <Sequell> 372/43712 games for bot (!@goodplayerspre20 t cv>=0.20-a): N=372/43712 (0.85%)
03:12:30 <Rotatell> Unstable branch on crawl.beRotato.org updated to: 0.21-a0-43-g1479571 (34)
03:33:12 <Kenran> Good morning everyone. Are strings of the form @some_text@ something that is replaced when the game is running?
03:34:42 <alexjurkiewicz> Kenran: in the source/dat directory? yes
03:34:54 <Kenran> Ah thanks
03:35:06 <alexjurkiewicz> see ../docs/develop/monster_speech.txt
03:36:01 <Kenran> I've gotten a ghost moth message (You feel you are....) ending with @player_only@ ingame. But searching Mantis has shown me this bug at least 5 times (being closed, resolved, and new from 2015).
03:36:44 <Kenran> So I shouldn't submit a new issue, right?
03:36:48 <Kenran> Since one is still "new".
03:36:58 <Kenran> (It's from 0.16 though)
03:39:06 <alexjurkiewicz> can you paste the line you received?
03:39:30 <alexjurkiewicz> was it: You feel you are being watched by something. @player_only@
03:39:32 <Kenran> "You feel you are being watched by something. @player_only@"
03:39:34 <Kenran> alexjurkiewicz: Yes
03:39:48 <Kenran> alexjurkiewicz: Someone wrote it happens when a ghost moth targets/hits a neutral target
03:40:12 <Kenran> alexjurkiewicz: In my case it seems like it might have happened with a ghost moth targeting a confused draconian though, as there were no summons or neutral monsters.
03:40:30 <alexjurkiewicz> sounds like there's a bug in the handling for @player_only@, since the docs say it should have been removed
03:41:05 <Kenran> alexjurkiewicz: When was it removed according to the docs, do you know that?
03:41:05 <alexjurkiewicz> you should post an update to the most recent open ticket with your reproduction, the scenario might help track down the exact issue
03:41:13 <Kenran> Okay thanks, going to do that!
03:41:19 <alexjurkiewicz> I mean the docs say the tag is cut from the line before the line is printed
03:41:27 <alexjurkiewicz> You can explicitly make
03:41:27 <alexjurkiewicz> a message player-centric by appending "@player_only@" to the end of any
03:41:27 <alexjurkiewicz> of the lines in the message, which will be removed before displaying it
03:41:28 <alexjurkiewicz> to the player.
03:42:15 <Kenran> https://crawl.develz.org/mantis/view.php?id=9201 is the most recent one I found (and the only still open one). I'm going to append my comment and screenshot there, even if this issue was for 0.16.
03:42:38 <alexjurkiewicz> yep, the affected versions can be updated
03:43:03 <Kenran> Great! I'm kind of new to Mantis :)
03:47:11 <Kenran> alexjurkiewicz: I have added a comment and the screenshot. I guess upgrading the branch is something a dev has to do.
04:15:53 <Eksell> stickyfingers (L10 DgIE) ASSERT((int)Buffer.size() == expanded_keys_left) in 'macro.cc' at line 544 failed. (D:10)
04:27:31 -!- amalloy is now known as amalloy_
04:28:25 -!- amalloy_ is now known as amalloy
05:37:05 -!- amalloy is now known as amalloy_
08:21:28 -!- jeefus is now known as jefus
08:28:15 <advil> Kenran: don't worry about the version / branch field in an issue, as long as your comment is clear.  it's really only interpretable with respect to the original report.
08:29:13 <advil> !crashlog
08:29:15 <Sequell> 17732. stickyfingers, XL10 DgIE, T:15462 (milestone): http://crawl.xtahua.com/crawl/morgue/stickyfingers/crash-stickyfingers-20170606-081533.txt
09:10:16 <Kenran> advil: good to know, thanks
09:15:04 -!- Fixer_ is now known as Fixer
09:15:40 <Kenran> It's kind of sad that CDO didn't participate in the tournament. It's the least laggy server for me by far.
09:17:42 <nikheizen> !blame napkin
09:17:42 <Sequell> I pronounce napkin... Guilty!
13:02:27 -!- amalloy_ is now known as amalloy
14:03:41 <Cheibriados> 03advil02 07* 0.21-a0-44-gb82119a: Fix up restoration/sustab innate muts on load 10(2 hours ago, 1 file, 26+ 0-) 13https://github.com/crawl/crawl/commit/b82119a2b70b
14:03:41 <Cheibriados> 03advil02 07* 0.21-a0-45-gc929894: Fix NOWIZARD builds 10(37 minutes ago, 5 files, 34+ 29-) 13https://github.com/crawl/crawl/commit/c929894b0a48
14:03:41 <Cheibriados> 03advil02 07* 0.21-a0-46-gdf2c162: Extremely basic lua bindings for fsim 10(35 minutes ago, 8 files, 114+ 1-) 13https://github.com/crawl/crawl/commit/df2c16248b20
14:05:07 <advil> ugh, checkwhite contamination in the fsim one
14:10:11 <amalloy> i am not a fan of mixing whitespace with functional changes, but even i would forgive such a small transgression as https://github.com/crawl/crawl/commit/df2c16248b20#diff-134a8114de41ab35f33c746461b4b0c3R2977
14:11:15 <advil> heh it's annoying because I was trying to make that commit somewhat cherry-pickable, and that adds another conflict to resolve
15:09:47 <Jorgrell> Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-46-gdf2c162 (34)
16:11:44 <advil> %git 85fd691b71e8e76133508bff88e2b1aa696e4d6a
16:11:44 <Cheibriados> 07Brannock02 * 0.20-a0-312-g85fd691: Fix damage modifiers at 9 and 11 Str (Roleplayer) 10(6 months ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/85fd691b71e8
16:11:47 <advil> %git 10ed074002f6b149313ed3068e4ce06e40b083a2
16:11:47 <Cheibriados> 07Brannock02 * 0.20-a0-313-g10ed074: Correct a value (|amethyst) 10(6 months ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/10ed074002f6
16:12:05 <advil> so these commits do seem to have slightly increased damage across the board
16:12:56 <advil> it's not huge, and my fsim thing is testing it in slightly weird circumstances (no lua binding for setting skills so), but it is noticeable in fsims with a lot of iterations
16:13:48 <minmay> yeah, they increased it a bit
16:16:15 <advil> no idea if it's enough to account for the winrate jump, I wouldn't have expected so, but if it's not that I'm pretty sure there's no other change that affects damage or accuracy like that in 0.20
16:16:16 <minmay> it was either that or decrease damage at 10 str
16:16:22 <advil> yeah, the changes seem right
16:17:00 <advil> not complaining about the commits. but we have been trying to figure out for the last few days why the winrate has slightly but noticeably increased this t
16:18:51 <minmay> an extra 2.5% base damage on average that gets rounded down is probably not the culprit
16:19:29 <advil> probably not. But by isolating that I think I can eliminate other problems with damage or accuracy.
16:24:48 <minmay> IIRC winrate is always higher earlier in the tournament, so it'll go down probably
16:27:34 <advil> hmm maybe read the backlog :)
16:27:36 <geekosaur> but the comparisons are against similar timeframes in earlier tournaments
16:28:26 <minmay> sry
16:28:33 <advil> np
16:31:21 <minmay> it looks like people are using more potions and scrolls?
16:31:32 <minmay> that has been going up for a while though
16:35:15 <CanOfWorms> we did increase the amount of hw/tp/haste when the wands were removed
16:35:54 <advil> yeah, that's come up...someone ran objstat yesterday I think...it's not as big of an average increase as you might expect
16:35:57 <advil> could be a factor though
17:38:53 <Floodkiller> question: should wand acquirement continue to exist on its own at this point, or could wands merge into misc?
17:39:07 <Floodkiller> a chance at scattershot/clouds/acid isn't really worth burning acquirement for, but could be an okay result when fishing for a good evocable anyway
18:16:09 <Lantell> Unstable branch on underhound.eu updated to: 0.21-a0-46-gdf2c162 (34)
18:24:44 <amalloy> Floodkiller: seems reasonable to me, as someone who is mostly out of touch with any recent changes
19:19:55 -!- Fixer_ is now known as Fixer
19:56:17 <gammafunk> !nchoice
19:56:20 <Sequell> Time for a new nchoice! It will appear shortly on the tournament website (if it hasn't yet). Type "=nemelex XXXX" to update !nchoice with the new combo, where XXXX should be replaced by the new combo.
19:56:35 <gammafunk> !kw nchoice
19:56:35 <Sequell> Keyword: nchoice => DEAM|DDAK|HOCj|BaEn|FoEn|MiVM|HaIE|VSEE|FeNe|GhWz|CeFE|SpTm|MuAM|TeWr|VpSk|KoAE|GrSu|OgEE|BaHu|MfCj|VpFE|DEAs|HOSu|BaAr
19:56:45 <gammafunk> =nemelex HaAE
19:56:49 <Sequell> Defined keyword: nchoice => DEAM|DDAK|HOCj|BaEn|FoEn|MiVM|HaIE|VSEE|FeNe|GhWz|CeFE|SpTm|MuAM|TeWr|VpSk|KoAE|GrSu|OgEE|BaHu|MfCj|VpFE|DEAs|HOSu|BaAr|HaAE
19:56:51 <gammafunk> =nemelex FoWz
19:56:54 <Sequell> Defined keyword: nchoice => DEAM|DDAK|HOCj|BaEn|FoEn|MiVM|HaIE|VSEE|FeNe|GhWz|CeFE|SpTm|MuAM|TeWr|VpSk|KoAE|GrSu|OgEE|BaHu|MfCj|VpFE|DEAs|HOSu|BaAr|HaAE|FoWz
19:56:57 <gammafunk> !nchoice
19:57:03 <Sequell> FoWz: 0 wins || beargit: CJR, L4 Magician of No God
20:33:19 <advil> !lm grimtooth
20:33:27 <Sequell> 14518. [2017-06-06 19:35:27] grimtooth the Impaler (L17 MfGl of Dithmenos) left the Spider Nest on turn 40630. (Spider:1)
20:34:37 <advil> ^dump grimtooth
20:34:37 <Rotatell> http://crawl.beRotato.org/crawl/morgue/grimtooth/grimtooth.txt
20:35:13 <advil> !lg grimtooth
20:35:15 <Sequell> 2096. grimtooth the Frost Mage (L14 MfIE of Hepliaklqana), slain by a komodo dragon on Lair:4 (spider_nest_entry_webs) on 2017-06-04 16:36:14, with 42219 points after 21910 turns and 4:10:41.
20:39:25 <gammafunk> advil: no further big leads on the mysterious winrate inflation problem?
20:39:46 <nikheizen> it's because i'm playing
20:40:02 <gammafunk> that would explain more if winrate were decreasing
20:40:31 <nikheizen> :3c
20:40:36 <advil> not really...there was a commit that increased the damage a bit across the board back in dec, but I don't think it's big enough.
20:41:10 <advil> I cross-checked fsim data between trunk and 0.19.1, and while damage is slightly higher now if measured right, that commit accounts for it all
20:41:10 <gammafunk> oh...actually
20:41:26 <gammafunk> there is a thing I want to do with mapstat/objstat
20:41:32 <gammafunk> let me go try that
20:41:36 <advil> %git 85fd691b71e8e76133508bff88e2b1aa696e4d6a
20:41:36 <Cheibriados> 07Brannock02 * 0.20-a0-312-g85fd691: Fix damage modifiers at 9 and 11 Str (Roleplayer) 10(6 months ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/85fd691b71e8
20:41:38 <advil> and
20:41:49 <advil> %git 10ed074002f6b149313ed3068e4ce06e40b083a2
20:41:49 <Cheibriados> 07Brannock02 * 0.20-a0-313-g10ed074: Correct a value (|amethyst) 10(6 months ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/10ed074002f6
20:42:04 <advil> those commits seem correct in any case
20:42:29 <gammafunk> hrm, I also notice they're quite early in the 0.20 cycle
20:42:34 <advil> yeah
20:52:05 <Doesnty> aren't wanderers supposed to be able to use their starting gear
20:54:11 <elliptic> hm, I don't understand that commit message
20:54:27 <elliptic> doesn't it alter damage modifiers at all values of str, not just 9 and 11?
20:55:14 <geekosaur> ?/remember much
20:55:15 <Sequell> No matches.
20:55:22 <geekosaur> ?/recall much
20:55:23 <Sequell> Matching entries (1): rare_messages[4]: You wake up in a daze, and can't recall much.
20:55:53 <elliptic> %git 6c060d3
20:55:53 <Cheibriados> 07Ada48202 * 0.19-a0-50-g6c060d3: Tightened the 0 bonus/penalty range for STR. 10(1 year, 1 month ago, 1 file, 4+ 4-) 13https://github.com/crawl/crawl/commit/6c060d3771dd
20:55:58 <elliptic> that one too
20:56:42 <elliptic> am I missing something or did both the first Brannock commit above and this one increase damage (very slightly) for almost all values of strength?
20:57:24 <advil> I haven't sen that one, but the damage modifier fix did, as far as I can tell
20:57:43 <advil> but it was by like 4% at most for the cases I was testing in fsim
20:58:10 <elliptic> probably not really a problem (the buff is extremely tiny) but it would be nice if commit messages weren't misleading about what they did
21:02:00 <elliptic> I guess the combination of both commits there was about a 4-5% increase to weapon damage for +0 weapons
21:02:27 <elliptic> but one of them was in 0.19 and the other was early in 0.20 as you observed, so it seems unlikely to be related to this
21:02:29 <advil> 6c060d3 was in 0.19 though
21:02:34 <advil> yeah
21:04:21 <advil> it does modify base damage though, so I guess it could be more impactful than it seems?
21:05:08 <elliptic> yeah, for +0 weapons with no additional additive boosts (might, elec, pain, etc) that multiplier gets applied to the entire damage
21:07:43 <advil> maybe next t we should try modifying damage on purpose to see if it actually does something like this :D
21:08:49 <advil> or even better, we could do a/b testing in trunk
21:10:50 <advil> 14:45:18 <Brannock> I'll just fix the bug and then we can stew further over formulas
21:10:57 <advil> I guess the stewing just got delayed a few months
21:19:55 <minmay> I'm encountering a bizarre display bug with archmagi + statue form
21:20:18 <minmay> casting it without archmagi I get 36 AC or whatever
21:20:37 <minmay> casting it with a robe of the archmagi on, it displays as 31 AC, but is actually 40 (and shows this the next time the display updates)
21:20:44 <minmay> &dump minmay
21:20:46 <Sequell> http://crawl.berotato.org/crawl/morgue/minmay/minmay.txt
21:21:35 <minmay> this does not seem to occur with other power-dependent-AC forms
21:26:18 <minmay> I suppose I should find some other body armour to see if the same thing happens
21:36:34 <minmay> advil: imo continue the small damage increases by making the calculation more precise, use 39000 instead of 39 and multiply the values inside the random2 by 1000
22:26:56 <minmay> fr hitting a death knight with scattershot makes it run out of yred piety and it cant pain mirror anymore
22:42:30 <minmay> why do clouds of translocational energy block LOS
23:03:35 -!- introsp3ctive1 is now known as introsp3ctive
23:44:53 -!- mibe_ is now known as mibe