Character sheet and Tavern thread

Congratulations!

As usual, if you made some Crawl fanart yourself, or know someone who did, then do not hesitate to contact me (either by email or by tavern message).

Many thanks to all the contributors, and may your inspirations run free!

This is a very inopportune moment, but it seems that CSZO vanished from the Internet this morning. The maintainer Amethyst opened an emergency ticket with the ISP a few hours ago; there has been no response yet.

For the time being, please use CAO or CBRO. If the downtime persists, we will try to host the tournament page elsewhere until it is resolved.

Update 2014-04-16 14:43 UTC: The ISP says they are having technical difficulty with their fiber connections to two upstream providers, and have been working on the issue since it started about eight hours ago. Several customers are experiencing problems, not just CSZO. They expect the network to be “back up and running soon”. -neil

Update 2014-04-16 20:55 UTC: The network is back up, but there is still some packet loss for the time being, so expect some lag until everything is fully resolved. -neil

Crawl Art page

Consider yourself a good Crawler if you recognise the content of one or two pictures at once. And a hopeless addict if you get all four. :)

It looks very spartan right now. I would be happy if, over time, more drawings and possibly fan fiction or other art find a home over there. If you want to recommend something for the Crawl Art page, please contact me (private forum message or email). In the long run, we may need to set up an editorial office, but for now this will do.

    Commit: #f2529f3

    Rune lock.

    You need a rune to go to D:15. Downstairs of D:14 could be recoloured
    or replaced by a portal (maybe more consistent with D:27 zot portal).

This is a mandatory Vow of Courage I banner. The idea is quite old and was pushed after the 0.13 tournament so as to have as much time for testing as possible. You can read discussions about it on a old dev wiki page, or in several forum threads: here (locked), here (locked), here and here.

If the rune lock is considered to be good enough to stay for the next release, there will definitely be follow-up changes, cosmetic ones (locked staircase appearance, dungeon appearance before and after the lock) and balancing ones (unique generation springs to mind).

The rules are available on the tournament website. Some rules and banners have been changed. As usual, clans can be set up and changed until one week into the tournament.

Updated 2012-10-11: CDO now has 0.11 webtiles. Thanks, Napkin!

You can read the full article here.

After finding out about his blog entry (via reddit), I contacted him and asked for permission to link it here. Not only was John fine with that, he decided to thank everyone by creating a new opening splash screen for the game that will appear in the upcoming release.

Read his message about the response to the article and see a large version of the new splash art here.

Some private closing remarks by me: When we started the Stone Soup fork, one crucial part was to encourage players to become active with the development. This has worked very well, ranging over speech, descriptions, maps/vaults, bug fixes and patches for sometimes elaborate new features. An area where contributions were, and still are, particularly rich is tiles. It is both a pleasure to thank John for his amazing work, and a shame to acknowledge that so much went unused. The arguments brought forward against his ideas could have been mine, or in some cases actually were mine. Further discussion on the matter should go the comments, ideally on John’s blog. Many thanks John, and farewell!

For quite a while it was realised that MD, HO and Mi are too close to each other than is good for well-differentiated variety. In other words, developers agreed that cutting a species of the three would be an improvement. This was not urgent; that’s why it isn’t announced on the 0.10 development plan or anywhere else. However, it is also not a big change, so when we got the details ready in a session or two, we commenced with the wipeout.

First, theme. Hill Orcs are sacrosanct by Beogh’s protection. Even if you don’t like the orc god, there is no way we would get rid of HO as a player species. Choosing between MD and Mi, the lot hit the former. The reasons are less forcing than with HO, but are reasons nonetheless: Mi has a monster equivalent whereas MD has not (the monster dwarves are deep dwarves, akin to deep elves being the monster elves). Mi has an ancient mythos behind it (Minos, Crete, the labyrinth) as well as a portal vault in the game. MD on the other hand is the archetypical Tolkien-style dwarf, which in my book is a severe demerit.

Second, aptitudes. Removing a species does not happen in empty space. It might have impact on nearby species, and it certainly did so here. A table of the old and new aptitudes for Mi, HO and old MD is at the end of this text. The summary is this: new Minotaurs approached MD in certain areas (e.g. Armour is 2, also Stabbing, Traps & Doors, Evocations). Hill Orcs got additional magical competence.

Third, to dispel a myth. We didn’t use winning rates, or number of games played in order to make a decision.

Fourth, history. Over the course of DCSS, the following species had to resign: HD, GE, El, Gn, OM (Hill Dwarf, Grey Elf, Elf, Gnome, Ogre Mage). Of these, OM merged with Og into a more interesting species; the others left for the same reason MD goes now: lack of differentiation. On the other hand, some species were added over the years: Vp, DD, Fe (Vampire, Deep Dwarf, Felid). 0.10 will probably feature Octopodes. None of these are defined by aptitudes alone (but note that new Ogre is an exception!), which makes for additional complexity but also diversity. Note that species tweaking goes beyond adding and removing: Draconians and Demonspawns were changed (and will probably see further changes), and with some luck and effort, 0.10 will see more interesting Demigods.

An empty entry means no change from the old version of the species.

        Arm Ddg Sth Stb Shd T&D   Inv Evo        HP  MP   Exp
--------------------------------------------------------------
Mi old   +1  +1  -2   0  +1  -1    -1  -3        +1  -2   140
Mi new   +2          -1  +2   0        -1                 130
MD       +3  -1  -3  -2  +2  +1     0  +1        +1  -1   130
HO old   +1  -2  -2  +2  +1   0     0   0        +1  -1   100
HO new                             +2                 0      

        Fgt SBl LBl M&F Axs Pla Stv  UC   Thr Slg Bws Crb
----------------------------------------------------------
Mi old   +2  +2  +2  +2  +2  +2  +2  +1    +1  +1  +1  +1
Mi new       +1                             0
MD       +2  +1   0  +2  +2  -1  -1   0    -1  -1  -2  +1
HO old   +2   0  +1  +1  +2  +1  -1  +1     0  -1  -1  -1
HO new                                                   

        Spc Coj Hex Cha Sum Nec Trl Trm   Fir Ice Air Ear Poi
--------------------------------------------------------------
Mi old   -3  -3  -3  -3  -3  -3  -3  -3    -3  -3  -3  -3  -3
Mi new           -4  -4              -2                -2
HO old   -3   0   0  -1   0   0  -2  -3     0   0  -2   0  -1
HO new   -2                                +1  -1                  +3 Int, -1 Str
MD       -3  -1  -2  -2  -2  -3  -2  -1    +2  -2  -2  +2  -2

Implementable voting

The winning player is theglow with 4934 points, who won 15 games, including an 8 streak! Runner up is mikee with 4653 points, who won 11 out of 13 games, including an 10 streak (actually 11 games in a row, but two wizards in the list). The competition was close, mikee might have won if his level 16 Spriggan Enchanter would not have died to an orb guardian in Zot:5. The third player to get more than 4000 points is jeanjacques with 4333 points.

Another highlight: 78291 played the fastest game, both in real time (2:32 hours) and turn count (19460), winning a Deep Dwarf of Elyvilon. He only killed 64 creatures himself, thus proving the power of pacifism.

The most successful clans were: The Adorable Catlobes with 14201 points, followed by wucad mu-tang clan aint nuthin ta fuck wit with 9074 points and Zignacio with 8276 points.

The ubiquitous numbers (in parentheses for August 2010):

Days:                       16         (31)
Players:                  1337         (1091)
Players getting a rune:    264 = 20%   (229 = 21%)
Winning players:           116 = 8.7%  (109 = 10.0%)
All games:               37593         (58428)
Non-quit games:          27049         (44884)
Winning games:             232 = 0.9%   (321 = 0.7%)
Won race/class combos:     128         (101)

Most everyone agreed that the new rules encouraged diversification. This can be seen when comparing the most popular winning combinations (three or more wins):

2011:
 8xSpEn  7xMDFi  7xMiBe  6xDDNe  6xDEWz  6xKoBe  6xVpAE  5xCeWz  5xHOAr  5xHaVM
 5xKoSk  5xMfIE  5xOgCj  5xTrAs  4xDECj  3xDsVM  3xGhMo  3xHECj  3xHEWz  3xMfGl
 3xNaTm 3xSpDK

2010:
49xSpEn 23xDDNe 22xMfCr 18xGhFE 18xHaTm 17xSEAs  9xMfIE  7xHEWz  7xHaBe  6xHECr
 5xDDCK  4xDECj  4xDSNe  4xHEIE  4xMfTm  4xSpVM  3xDEWz  3xDSAs  3xDSBe  3xDSVM
 3xHEFE  3xKoCr  3xMDEn  3xSEIE  3xSpSt

USAGE AND WEBTILES:

Server usage jumped up (about 400 players/day), not least because tiles players were able to join in the fun – a feature that’s been long asked for: Webtiles, Usage

PLAYER HIGHLIGHTS:

Winner was mikee (4191 points), narrowly edging out elliptic (4107 points) by managing the last win in the tourney (which gives bonus points).

mikee started with a two streak (including the first 15-rune win of the tourney), followed by two duds and then a seven streak! (KeAr^A, TrCr^E, MiAr^O, SEIE^K, HuVM^V, HOHu^T, HECj^S)

theglow played just five games and won them all! (SEWz^V, HuSt^S, GhGl^M, CeTm^O, DEIE^A)

casmith789 had the most combo high scores (22).

Players could go for some additional goals (Pennants). Only two got them all: 78291 and mikee.

CLAN HIGHLIGHTS:

Winner was Zinja, who was also the only clan to win at least one game with each species.

Second place was narrowly won by catonkeyboard (the team with the most combo high scores) over killdudes (aptly enough, the team with the most uniques killed… only Tiamat escaped them).

INTERESTING STATISTICS:
(brackets are about last August’s tournament, which took a full month)

number of players: .......................... 1523 (1091)
number of players with at least one win: ..... 108 (109)
number of wins: .............................. 233 (321)
average number of wins per day: ............. 14.5 (10.3)
distinct winning combinations: ............... 149 (101)*
overall winrate, quits excluded: ........... 0.79% (0.72%)
* Rules have been changed to make niche combinations more attractive.


Thanks to everyone who helped with the infrastructure and cheers to participants!

• Provide (some) divinations, as the spell school got axed in 0.7.
• An introverted/laidback god: many deities ask for a lot attention by the player, e.g. praying, sacrificing, training Invocations. There was demand for a god who is not like that.

The current incarnation meets both goals. Here is a list of properties:

• Piety is gained by exploration. Piety gain is considerably faster if you wear some cursed items. (Remove Curse scrolls can be turned into Curse Foo scrolls. There are now also Curse Jewellery scrolls.) Piety drops over time.
• All abilities but two are passive and focus on detection. You will map your surroundings, see presence of nearby items and monsters, and identify items.
• One active ability is scrying: this turns all walls into glass to look though, but only for you! So you can have a look into some vault before deciding whether it is worth or reasonable to storm it.
• The other active ability is about reskilling. You can sacrifice one skill in order to boost another. The interface for this should be very straightforward, thanks to a completely reworked skills screen. Needless to say, neither ability can be abused (e.g. you cannot smite through scryed walls, and you won’t produce virtual xp by reskilling cross-trained skills.)

What is the theme of such a god? The question is not at all rhetorical, there have been many complaints in advance about the concept.

• Ashenzari seems very unflexible when it comes to the actual world. Presumably the deity is cursed in some way as well. Hence, followers can demonstrate dedication by donning cursed garb and gear.
• On the other hand, the god is one of knowledge. Pious followers can tap some of this power so as to see more than their neighbours will.
• The pinnacle of outward torpor yet inward flexibility is reskilling. While Ashenzari cannot change items or threaten monsters for you, there is help to change your way of life to accomodate to new situations.

What does it boil down to in actual play? I am happy to have won a SpAs of Ashenzari. Here are my impressions:

• There is a curse mini-game. Scrolls of Remove Curse and Curse Foo are valuable, throughout the game. I found this interesting in all of my testing games. An unfortunate Demonspawn happened to have only two rings early on, Hunger and Sustenance; so I cursed both.
• My winning Spriggan used only cursed jewellery. This was enough to get all the piety I needed — an indication that the piety boosts are too good at the moment.
• Whatever you curse, it will make you more inflexible. This is also interesting in practise.
• Seeing the presence of monsters is often very useful, when stabbing, fleeing, etc.
• You will use the two active abilities not very often but Scrying can be of high tactical value. It is not very expensive, too. Reskilling gives great strategical flexibility. If you are a caster, you don’t need to sweat about choosing schools to specialise. It would be possible to turn all your skills into Fighting and Spellcasting (something that was mentioned as an argument against the mechanic). I believe that this is a shortcut to kill off your character.

So the god is perfect then? Of course not. I think the current state is playable and (naturally) diverse enough, but there are rough edges:

• In order to gain the curses bonus for piety gain, you want more than half of your potential armour items cursed. Or at least two jewellery pieces. Or your weapon. In any case, there is no benefit for cursing more than that. But there should be. One solution would be to give a little piety boost for cursedness as now, and a higher boost for full cursedness (armour and jewellery only).
• Item and monster detection only tells you that something’s there. While there is Scrying, it would be much better to give slightly better feedback, e.g. to announce monsters in three levels (puny, approximately in depth, OOD) and to distinguish between plain and rare items.

As ever, you can play Ashenzari on CDO and, as usual, feedback is welcome. We are finetuning the god, so don’t expect your “This is a dumb god” to have a far-reaching effect.

Click here for the official tournament web page,

Some interesting numbers (those in brackets refer to the 2009 and 2008 tournaments):

Games played:       44884  (28306, 13771)    [excluding quitters]
Number of players:   1091  (737, 385)
Number of winners:    109  (90, 50)
Numbers of wins:      321  (286, 169)

(For comparison, the Nethack devnull 2009 tournament had 11910 games by 354 players with 292 ascensions. These numbers seem suspiciously low, so here are the 2008 figures as well: 22812 games by 1305 players with 300 ascensions.)

The winner of the Crawl tournament 2010 is elliptic with 4121 points and 17 wins. The runners-up are Jaeger (3351 points, 17 wins) and 78291 (3274 points, 10 wins).

The best clan is Team Kiku, they dominated the competition.

Some fancy records achieved during the tournament:
Longest streak: elliptic with ten wins in a row!
Fastest win: qwqw won a SpEn in 14318 turns!
Quickest win: 78291 won another SpEn in less than 72 minutes!

For the next year, it may be desirable to have another North American server. The load was quite heavy during the first week.

Apart from that, there is talk of shortening the tournament to cover only 16 days. For comments on that, and also on the scoring, please comment here.

Congratulations to all winners and thanks to everyone involved, including all players and the good souls announcing this event on boards and forums!

The Crawl devteam.

The consolidation aspect of 0.7:
* Monster double actions due to energy randomisation always include a move. The double action was a source of a lot of controversy. In 0.7, an adjacent monster won’t hit you twice, although a monster a step away might still approach and hit you in a single turn.
* Make heavy armour guaranteed damage reduction percentage independent of armour skill. This makes heavier armours a bit better again, after the severe but long overdue nerf in 0.6.

In other news, there have been a lot of improvements to Tiles, Tutorials (tutorial and hints mode are now distinct) and the interface in general. Crawl will also officially feature Chapayev’s awesome Sprint variant.

Gameplay-wise, the biggest changes are these:
* Demonspawns will feature less lottery; all mutations grow to the full; also, mutations are meaningful, in particular, many of them are specific to the species.
* Zin has finally seen the light help: Recite is less risky, hence more useful. Also, there’s a new ability, which allows to imprison intelligent monsters.

Feedback on DS, Fi and Zinners are appreciated. We won’t re-design these on short notice, with release being so close. But comments on how far you get (and got in older versions) would help to see gauge we stand with the new AC/DS/Zin.

Some more changes which will be of interest:
* Aptitudes now on a scale +5 to -5.
* Remove Death Knights and Thieves.
* Show effective weapon speed in ‘@’ character info.
* Walls in the Slime Pits are acidic.
* Changed shop distribution between branches.
* Removed instant stat death.

Happy Birthday, Stone Soup!

The distinction between ASCII and Tiles use it particularly interesting. Thanks, everyone, for your vocal participation!

In other news, the DOS version has long been available for download from the Sourceforge site. And for Mac users facing problems when installing 0.5 we’ve added some documentation to hopefully help with that.

Edit: Tiles port and Mac OS X binaries are available now.