Angband Forums Greater Vault Frequency
 Register FAQ Members List Calendar Search Today's Posts Mark Forums Read

 September 17, 2009, 07:13 #1 d_m Angband Devteam member     Join Date: Aug 2008 Location: Philadelphia, PA, USA Age: 38 Posts: 1,516 Greater Vault Frequency I just wanted to share some observations about room frequency, etc. to see if they seem consistent with people's observations. The following chart breaks down probabilities that the first room of a dungeon is of a particular type by percentage (so 1.2 is 1.2%): http://www.bearhome.net/erik/room-gen.txt. In theory these percentages work for each of the 50 rooms being tried, but in practice the probabilities shift due to lack of space for larger rooms (particularly greater vaults). I computed these numbers by reading the code and doing a little bit of basic statistics. Due to the complexity of measuring the probability that a particular level would have a particular room type, I decided to resort to dynamic analysis. After collecting a bunch of data via debug mode it seems like the probability for a particular level to have a greater vault is roughly equal to the percentage chance that the first room will be a greater vault at that depth times 10. That is: Code: ```0: 0% 10: <1% 20: 1% 30: 2% 40: 4% 50: 6% 60: 9% 70: 12% 80: 16% 90: 20% 100: 25%``` This might be a little generous... using x8 or x9 as a multiplier might be a slightly better fit but x10 was easy to compute for this post. Does this square with people's general experience or beliefs? Fair warning: I'm not necessarily committed to attempting to maintain the exact distribution of all room types in possible generation refactors... but since greater vaults are an important/fun part of the game, it seemed like a good thing to measure before and after. __________________ linux->xterm->screen->pmacs Last edited by d_m; September 17, 2009 at 07:18.
 September 17, 2009, 07:58 #2 Pete Mack Prophet   Join Date: Apr 2007 Location: Seattle, WA Posts: 5,014 Donated: \$40 This is a case where bootstrap simulation will trump calculations, any day...
September 17, 2009, 19:24   #3
PowerDiver
Prophet

Join Date: Mar 2008
Posts: 2,712
Quote:
 Originally Posted by d_m After collecting a bunch of data via debug mode it seems like the probability for a particular level to have a greater vault is roughly equal to the percentage chance that the first room will be a greater vault at that depth times 10.
I would have guessed much less than 10x, maybe 6x but surely your stats are better than my guesses.

OTOH, vaults have different sizes. Does that matter in the placement? If so, seeing how much it changes for the largest vaults would be of interest.

Were your experiments on a fixed dungeon size, or the random sizes currently in 3.1.2? It would not surprise me for significantly different behavior on an 80%x80% level.

September 18, 2009, 03:13   #4
d_m
Angband Devteam member

Join Date: Aug 2008
Age: 38
Posts: 1,516
Quote:
 Originally Posted by PowerDiver I would have guessed much less than 10x, maybe 6x but surely your stats are better than my guesses. OTOH, vaults have different sizes. Does that matter in the placement? If so, seeing how much it changes for the largest vaults would be of interest. Were your experiments on a fixed dungeon size, or the random sizes currently in 3.1.2? It would not surprise me for significantly different behavior on an 80%x80% level.
I'm pretty sure that no matter which vault will end up being placed, it checks to make sure there is enough space to accommodate the largest vault. I did these experiments on a full-size dungeon, hard-coding the scaling factor to 100%. I think you're right that smaller dungeons will get greater vaults much less frequently.
__________________
linux->xterm->screen->pmacs

 Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 Thread Tools Display Modes Linear Mode

 Posting Rules You may not post new threads You may not post replies You may not post attachments You may not edit your posts BB code is On Smilies are On [IMG] code is On HTML code is Off Forum Rules
 Forum Jump User Control Panel Private Messages Subscriptions Who's Online Search Forums Forums Home Angband     AAR     Vanilla     Development     ToME     Sil     Variants     Competition The real world     Idle chatter     Oook! Obsolete     v4

 Similar Threads Thread Thread Starter Forum Replies Last Post bebo Vanilla 4 January 27, 2009 16:50 Pete Mack Vanilla 11 January 19, 2009 21:34 vorondil Vanilla 12 December 16, 2008 18:48 xxitheworld Variants 4 February 18, 2008 05:46 Dragonboneman AAR 30 November 2, 2007 20:27

All times are GMT +1. The time now is 18:03.