I’ve been doing BWL transmog runs for the past few months and have convinced myself that there is an issue with drop rates on some items off of Chromaggus’s loot table. I decided a few weeks ago to start documenting all the drops for all my characters, which I will include in this thread simply for documentation’s sake. I understand that “RNG is RNG” but it certainly feels like the drop rates on some items are lower than they should be (for “simple” epics).
(Updated 4/9/2019)
For the period of 12/18/18 - 4/8/19:
Runs completed: 204
Total Items Dropped (ex. Whistle): 816
Items Not Yet Seen:
Yes, completely understand - though this isn’t a case of not seeing an item on a character that necessarily needs it, it is a case of never seeing these items drop at all. And the stats above are only when I started tracking, as I have been doing these runs for the last few months. So ~150 runs without seeing any of these items drop seems a little crazy to me.
Again, just going to be documenting all my runs here to show that the drop rates on some of the items appear to be lower than what one may expect.
I remember in vanilla I had to wait over 2 months to even see taught dragonhide belt drop from one of the drakes (it could drop from all three) and we didn’t see a claw or chromatically tempered sword for over three months. The item drop rates were quite low on some items (~5%). Couple that with a large group and it wasn’t easy to get the specific items you wanted.
Yes BWL is famous for this
youre right maybe there is a bug ?
but normally its just that there are some items that drop 20-30% and some are like 2%
Nefarian has some items that are like 1% drop and some that are 30-50%. that’s how it used too be in vanilla.
Yep! So part of this is me just documenting it, secretly hoping it comes across someone’s desk that is like “hmm, does it really make sense that Turoken and his alt’s have run BWL 500 times and sitll not seen some of the items, maybe we should look into that.”
May be a bug, may be intentionally really low drop rates, not sure…but we shall see.
i have run BWL countless times, easily in the hundreds. i have seen ashjrethul drop two times. you might just be getting a case of the “it’ll never drop” but it does in fact drop. though, the drop rates on old items are seemingly impossible, i will say, they do in fact drop. whether or not blizz wants to change the drop chances is a whole nother story. in which case you should say “please equilize drop rates on old items” rather than “it’s not dropping enough for me”
the claw drops plenty of times i’ve seen it lots. everything else drops normally but the crossbow in my runs has been like seeing Sasquatch sometimes.
Yes, I think “please equalize drop rates” is specifically what I am calling for…and so the purpose of this post (and your replies) is really to show that they really do not appear to be equalized at all.
(and also to demonstrate my own really bad luck, I suppose.)
Well, I think we are talking a little bit more than “rare” at this point. For 7 items (out of 22) to never drop on 71 runs where 4 items drop each run, we are seeing numbers where any statistics major would likely reject the null hypothesis that his drop rates follows something in line with what WoWHead shows historically. I’m no statistics major, but I’ve done the math, and it checks out.
I am going to continue documenting and editing the above, but it certainly feels like something is wrong here.
so let’s look at some examples. claw of chromaggus wowhead says drops 6% of the time. so we know that this is 6% per boss kill, because if it was 6% per item received then boss drops would have to total 100% which they don’t.
so over 71 boss kills, 6% chance per kill, your odds of having seen this are about 98.7%.
elementium threaded cloak has a 12% drop chance per kill, 71 boss kills, your odds of having seen it are 99.98%, so it is 1 in 5000 that you wouldn’t have seen this yet.
the rest of the items are in the 10-12% drop rate range from what i’ve checked. i wouldn’t call this degree of bad luck conclusive. it is possible that you’re just very unlucky. but given all 7 items having a low (~1-5000) chance of having not seen them yet, i think it is pretty likely that wowhead’s estimated drop rates are not correct. whether that is a bug or not, we can’t really say.
you have no idea what you’re talking about. if you care to learn i can explain but i’ve tried to explain before on these forums and am met with a stone wall of willful ignorance every time.
I am still not sure your math is correct if eveey item had an equal chance of dropping that would be 1/23 roughly 4 percent, 23 items on loot table you are grabbing 1 specific piece. Succesive attepms are idenpendent of eachother. I dont see how you think at the end of 71 runs of a 6 percent to drop item you have a probability of 98.
I need to see your math, it still looks like youbare saying that succesive runs effects you probability of seeeing an item.
My assertion that something feels off is based on performing a chi-square “goodness of fit” test. If we assume that the distribution of drops follows the rates from WoWHead, then we would see the following in terms of expected number of items based on my 71 runs (along with what I actually saw):
Item
Actual
Expected
Chi Contribution
Bloodfang Spaulders
13
18
1.43
Dragonstalker’s Spaulders
15
17
0.31
Pauldrons of Transcendence
26
17
4.53
Chromatic Boots
16
17
0.06
Epaulets of Ten Storms
27
17
5.94
Stormrage Pauldrons
16
17
0.06
Nemesis Spaulders
16
16
0.00
Netherwind Mantle
16
16
0.00
Taut Dragonhide Shoulderpads
23
16
3.13
Empowered Leggings
21
15
2.32
Judgment Spaulders
20
15
1.68
Pauldrons of Wrath
16
15
0.11
Angelista’s Grasp
20
14
2.49
Chromatically Tempered Sword
15
11
1.35
Elementium Reinforced Bulwark
24
10
18.06
Taut Dragonhide Gloves
0
10
9.60
Elementium Threaded Cloak
0
9
8.89
Girdle of the Fallen Crusader
0
8
8.04
Primalist’s Linked Waistguard
0
9
8.57
Shimmering Geta
0
8
7.63
Ashjre’thul, Crossbow of Smiting
0
5
4.58
Claw of Chromaggus
0
4
4.49
The “chi contribution” is found by taking (actual - expected)^2 / expected. When we add this up we get a value of 93.28. Checking this against a chi-square distribution with 21 degrees of freedom (22 items - 1) gives us a p-value that is essentially 0…meaning that the likelihood of seeing these results is near 0.
Now of course, all this says is that we can reject with a high-level of certainty our original assumption that the drop rates follow those historically recorded on WowHead.
Also, as I said above, the other part that makes this feel even more than just “really, really, really bad luck” is that the 71 runs are only those that I recorded, I had been running the dungeon for a few months before when I started noticing this trend and starting recording. So it’s been somewhere on the order of 200 runs with my not seeing these items drop.
Interesting turoken - I haven’t done that sort of analysis since college. So I was guessing that the wowhead values were wrong but it wasn’t conclusive, but i think it is actually off the charts conclusive that the wowhead values are wrong. That unfortunately still leaves us unsure whether they are bugged. You might make a post in bug reports though.
@baltysalls - let’s look at one item. An item that has a 6% drop rate. We know that to combine the probabilities of each independent random event, we multiply the odds of each event together. If you don’t understand this, reread my link. So what we are trying to find here is the chance of 71 runs where the item does NOT drop. So the odds of it not dropping on any given run are 94%. So, what we need to do is multiply .94 (odds of no drop) by itself 71 times. Or .94^71. this gives us .012. so there is a 1.2% chance of completing 71 runs without seeing the item, or alternately a 98.8% chance that we will see it.
Well, I don’t think he is asserting that it is “100% right,” however, the law of large numbers ensures that a large enough sample should produce something close to the expected value - and I think the ~18K observations certainly qualifies as “large enough.”