With the release of the new patch this week I wanted to get a better understanding of the drop rate of items, and how and if magic find affected it.
After getting 5 NV stacks I began taking screenshots right after an elite kill, and of every identified item. I spent around 11 hours (with a few breaks) across wed-fri running Act 3 on Inferno by myself and with 1 or 2 friends.
After getting 5 NV stacks I began taking screenshots right after an elite kill, and of every identified item. I spent around 11 hours (with a few breaks) across wed-fri running Act 3 on Inferno by myself and with 1 or 2 friends.
We killed elites all across the Act, and all had similar magic find gear, and would swap it out every time we could before the kill, being unable to do so only a few times across the 111 kills I tracked. Across these kills I recorded receiving 232 rare (yellow) items, and have recorded their details here:
https://docs.google.com/spreadsheet/ccc?key=0AjZShULpMOMwdGlPVTBtcGRwNTlsMDZPMGp6M003cGc
Some notes about the data: there were 2 instances I believe where my machine took the screenshot showing all items on the ground, but for whatever reason did not get screenshots of the identified items, I have indicated their level and name in the spreadsheet, but do not recall the exact stats, other than remembering that they fit the mold of the rest of the drops (lousy). Also for weapons I did not record elemental damage since it is factored in to the total DPS.
High level Points:
High level Points:
- Sample size is not enormous, but is reasonably sized so trends should hopefully be apparent
- Unfortunately I don't have control data to compare against prior to 1.03
- Item Levels of the loot were pretty close to Blizzard's advertised rates
- Going from 101 to 338 Magic Find only minimally improved the quantity of rares received per elite kill (0.39 more rares per elite)
- Magic Find did not seem to affect the average item level of items received (should it? I was never clear on this point)
- Nearly all of the items I found were bad and would sell for very little on the AH. Critics would say this is a sequence of bad random rolls, but after this many kills and 11 hours playing I would expect to see at least a few exceptional items
- Weapons rolled very low DPS, 1-handers maxed at 794 DPS, the average DPS was 316.8, 2-handers maxed at 914 with an average of 442 DPS. A quick AH survey shows 1-handers capping at nearly 1300 DPS and 2-handers at 1553 DPS. I did not track blues, but I did look at them and they never had higher DPS than the rares I received.
- Armor was exceptionally bad, I didn't get a single item that had a primary stat, vitality, and all resist on it
More analysis is on the Analysis sheet of the above link, I tried pasting here and the formatting just didn't work out.
Suggestions:
- Be more open with exactly how magic find works, at least currently it does not seem to increase the rare quantity by the amount it claims to, and does not affect item level (based on the sample size I have)
- Higher item level items should not be able to roll really low stat values that match item levels in the 50s.. give them a minimum value stat roll and an improved chance of rolling 'good' stat combinations (seehttp://img651.imageshack.us/img651/2/screenshot303z.jpg for an example of what is not useful, and actually pretty insulting on an ilevel 63 item)
- (Misc) Bosses should give an NV stack, or at the very least refresh the stack's timer
- (Misc) Please don't rush patches out the door without code reviews and regression tests on critical sections such as loot drop rates and stat rolls. Your customers have and are spending a lot of their precious leisure hours chasing after items that don't seem to be able to drop.
Thanks for reading!
