Changes between Version 1 and Version 2 of StatsHowTo


Ignore:
Timestamp:
03/20/11 16:30:19 (8 years ago)
Author:
fizzix
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • StatsHowTo

    v1 v2  
    21217: A final crawl through the dungeon is done and the items from uniques are recorded. 
    2222 
     23Generally the items that are tracked cover the useful items in the game.  However, there really is no limit as to what items could be tracked, and it might be worthwhile to just track everything. 
     24 
    2325== Diving vs. Clearing == 
    2426The ''Clearing'' option is used to simulate a game.  Every dungeon level from 1 to 100 is cleared, and all items are recorded.  After 100 levels have been completed, artifacts are reset to unseen and uniques are revived, and the simulation repeats for the number of iterations.  This is useful for getting relative artifact frequencies (comparing randarts to standarts, or evaluating changes in artifact frequency).  Along with the normal output, the ''Clearing'' option also has additional output for the first time certain items/effects are found in the dungeon, along with standard deviations. 
    2527 
    26 Diving is a little more simplistic.  It clears every 5th level N times, never marking artifacts as unseen or uniques as killed.  This simulation is good for getting relative frequencies of items or estimates for how likely out of depth monsters are discovered.  It will also help with basic frequencies, and total gold values.  It should not be used for any information regarding artifacts.  
     28''Diving'' is a little more simplistic.  It clears every 5th level N times, never marking artifacts as unseen or uniques as killed.  This simulation is good for getting relative frequencies of items or estimates for how likely out of depth monsters are discovered.  It will also help with basic frequencies, and total gold values.  It should not be used for any information regarding artifacts.  
    2729 
     30== Note on Randart Regeneration == 
     31Currently randart regeneration is broken.  The randarts are regenerated off of the previous randart set instead of the standart set.  As such, any systematic variation in artifact power will be compounded through each successive iteration. 
    2832 
     33== Drawbacks and Problems == 
     34Currently there are a very large number of problems in mimicking normal Angband play.  These are outlined below. 
     35 
     361. The player kills all monsters, even those that would not be fightable at the current depth. 
     372. Killing all monsters is not feasible.  Especially without generating extra monsters on the level. 
     383. Summoners like Qs will generate no loot, even though a draconic Q will generally give you a dragon with good loot. 
     394. There is no way of tracking the loot that comes from pits.  This is problematic, since it is a huge part of gameplay. 
     405. There is no way of tracking frequencies of vaults/pits/labyrinths/caverns and other macroscopic dungeon features.