Probabilities of item generation, by DL Google spreadsheet (2d table) Google spreadsheet (long table) (both contain expected count in 10000 items) DescriptionUnidentified individual generations(castings) and idenfied final results Particular books are not included in Identified, as each one is picked as separate item, and over 8K items game gets problems with performance(it was picking about 2 books/sec). Corpses(and lot of other food) rotted during generating/picking/dropping, so their number in identified section is reduced. Semi-raw dataFrom over 30,000,000 (10K per file) generations, splitted among different danger levels, troll wizard lvl1 in 1.2p9:* generations - unidentified * "heaped" -identified * normalized (prefix/suffix/parameters removed) Naming scheme: * irdl<X> := DL is probably X (calculated through maximum gold drop/44, later it was statistically tested that max/44.5 is more accurate) * sod<X> := number of blessed ?ofDanger read Artifacts frequencyMy experiments suggest that chance for item generated being artifact is 1/1000: Non-artifact items that were never generated
Empirical Danger Levels of itemsGoogle spreadsheetExtract from data "Semi-raw data" section.Not all artifacts are available, 4 items are unclear(it's written in comments colum) Empirical Danger Levels of monstersGoogle spreadsheet |