Talk:Obsidian Shard
Necessary?[edit]
If you already put up a link to creatures that drop obsidian shards, is it really necessary to sum a(n incomplete) list of creatures in FoW that drop these? Class 18:13, 24 December 2009 (UTC)
Night fall version of shadow army question.[edit]
Do they also drop obby shards and rubies and sapphires?
- For Obsidian Shards, no. Nothing outside of FoW drops them. For rubies/sapphires it's probably the normal everything-has-a-chance-to-drop-them chance. --Valshia 20:40, 10 February 2011 (UTC)
- If you read the article you will notice: "Random drop from specific creatures in The Fissure of Woe." [1] --Silver Edge 06:48, 13 February 2011 (UTC)
- Not everything can drop rubies and sapphires, see Category talk:Drops ruby#Not all monsters drop rubies. --Silver Edge 06:48, 13 February 2011 (UTC)
- I stand corrected then. --Valshia 18:16, 13 February 2011 (UTC)
- Not everything can drop rubies and sapphires, see Category talk:Drops ruby#Not all monsters drop rubies. --Silver Edge 06:48, 13 February 2011 (UTC)
Drop Data[edit]
Here's some data I gathered during my spirit-spammer runs through the spider cave (solo runs, normal mode). It's inconclusive, but maybe someone else is interested in extending it.
Monster | Drops | Kills | Expected drop rate | interval (95% confidence) | interval (99% confidence) |
Snarling Driftwood | 5 | 210 | 2.38% | 0.94% - 5.48% | 0.61% - 6.91% |
Armored Cave Spider | 24 | 714 | 3.36% | 2.24% - 4.97% | 1.96% - 5.60% |
Shadow Beast | 1 | 18 | 5.56% | 0.28% - 27.78% | 0.06% - 36.11% |
Haven't seen a shard from the monsters that aren't listed as dropping shards, either. Didn't bother counting Shadow Army creatures, because I rarely killed any.
The confidence interval is the interval of all possible drop rates that I cannot reject with p<0.05 (p<0.01). That doesn't mean that "it's 95% (99%) likely that the actual drop rate is in that interval", it just means that "for any drop rate outside that interval, my measured drops are less than 5% (1%) likely to happen". Not sure how mathematically sound that approach is, but it shows that even 700 kills aren't enough to be conclusive.
Bonus Points: deduce from the kill counts how many runs I failed or abandoned, and where :p Tub 10:25, 31 May 2011 (UTC)
Here's some more data from Shadow Rangers / Abyssals. Unfortunately I did not track them independently but it's a pretty large data set:
Monster | Drops | Kills | Expected drop rate | interval (95% confidence) | interval (99% confidence) |
Shadow Ranger / Abyssal | 137 | 4,350 | 3.15% | 2.63% - 3.67% | 2.36% - 3.94% |