Does the Twinhead database reflect ingame drop rates?

[jorim]

I'll start off with an example: The Eye of Sulfuras. On the twinhead database this is given a 1.34% drop chance from Ragnaros.

If I look at data from Allakhazam they put it at 3.70% drop chance derived from over 40k samples.

So it seems that it would be more correct to put the eye at 3% similar to the bindings from garr and geddon.

I think especially for very rare items that it is important that the drop rates are some what correct. I don't mind them being more rare than retail, but they certainly shouldn't have higher drop rates.

I don't mind doing extensive research to see which drop rates vary greatly from allakhazam/thottbot figures, and I also plan to check all rare mobs for their loot. But before I do that I want to know:

- Does the Twinhead database reflect the exact ingame drop rates? If not, why not?

- Regardless from the sample size, what errors can I expect in the allakhazam data? Does anyone know if there was any discussion in the past about all drop rates on allakhazam being a little higher or lower compared to reality? Caused for example by errors in the addon where it was not tracking loot correctly or tracked same data twice.

- Where did Kronos get his current data?

I want to avoid making unnecessary bug reports by making wrong assumptions.

Thanks in advance

 

[Rolexsweep]

I want to avoid making unnecessary bug reports by making wrong assumptions.

Well, you can still cross-reference the two datasets regardless of the methodology behind it (if you have access to this). That will atleast reveal the major discrepancies, but of course it doesn't really answer which of the two is more correct, or if Twinhead reflects ingame droprates at all.

 

[Cyndan]

...

I hope that 1.4% will stay as the droprate(or even slightly lower). You'll have to remember that retail had over 7 million subscribers in the month of the link you posted, and 1500 drops means that .0002% of the playerbase received that drop. If it were 3.7% here, with only a thousand (or few thousand at best) people, there would be Sulfuras' running around all over the place after a little while, making a legendary seem just like the gear you can get in retail now. Even though 1500 seems like a lot, I remember that I rarely ever saw one when I played Vanilla, and when I did it was a huge deal. I hope that they keep the droprate low, maybe even lower than 1.4% because it should be a big accomplishment when someone is running around with one.

Again, this is just my opinion and does not answer any of the questions you asked before people get all in a tizzy about it lol.

 

[Gourmandises]

- Does the Twinhead database reflect the exact ingame drop rates? If not, why not?
I bet the database reflects the set drop rates of the kronos server. The kronos peeps set up the database site and develop the game at the same time. I do not see any reason for them to set specific drop rates in the game then then post intentionally non-matching "estimates" on the same site they moderate.Unlike allakhazam and thottbot, who are not blizzard, do not know exact drop rates and do not have any power to set drop rates. The real question, I guess, would be if the twinhead database is accurate to real vanilla wow. If that was your initial question I would say no, as old databases still estimate drop rates.
Reflect estimated ingame drop rates from back when? Sure, unless they're not finished with that part of development process. In any case it does not hurt to express your concerns via the bug tool or whatever(unless explicitly notified otherwise).

- Regardless from the sample size, what errors can I expect in the allakhazam data? Does anyone know if there was any discussion in the past about all drop rates on allakhazam being a little higher or lower compared to reality? Caused for example by errors in the addon where it was not tracking loot correctly or tracked same data twice.
I would think that data errors can't be measured. If they could then we would compensate for the measurable error and be a ginormous amount closer to knowing the exact percentages, if not know it. If the errors could be detected on allakhazam surely they would have just removed invalid and borked submissions and we'd have the accurate info. However, most people do apply a give or take approach to things.

- Where did Kronos get his current data?
I dunno! Maybe they have cool archived versions of database sites from vanilla. Nobody's to know if blizzard's really changed any percentages from all old stuff but that doesn't mean you can trust things like live wowhead.

All in all, I really don't know anything