That stinks. I usually have that kind of luck but DF has been good to me, I’m getting at least one item per run. Knocks furiously on wood
Or just make it so if you trade an item you don’t lose the mog unlock.
Personal loot still feels better, but these are good changes.
Me to! There was a reason the greed/need system was replaced. Brining it back was an idiot move. It still has all the same problems as it did before. It needs to be permanently deleted!
That’s patently and provably untrue. Personal loot already had item weighting going on, just like Need/Greed. They literally played around with item weights multiple times during Shadowlands.
Try. In any system where the person is picked before the item (as PL was), you’re going to see the drop rate of every item be proportional to the portion of the raid who loot spec for that item.
I’ll even quote you the math I did earlier in another thread:
And before you can say “they can tune drop rates for certain items”, remember that the drop rate of any item is a function of the group comp. Tuning it for 1 comp will throw it completely out of whack for another.
2 Likes
people keep saying this but in my experience with 2 characters doing every wing available since LFR came out, I have never seen anything come close to TOXIC. Unless it’s all being done in whispers, it just ain’t true. In my experience.
2 Likes
Anyone who claims this has never been rage whispered for an item from PL that they didn’t want to trade.
There are players who will invent loot drama no matter what system is used.
2 Likes
I feel like those auto spam addon’s that would ask for loot as soon as you got it haven’t been in use, or maybe I’m just lucky in groups, but it GL stopped those it’d be worth it for that alone.
Nothing you’ve said is atypical behaviour for Group or Personal loot.
We even know for Group loot, we’ve had the history from the last time it was in the game that we had things like “Leather drops more because there are more leather-wearing classes in the game”, “Vanq token drops more because there are more classes using Vanq now”, 2h Agi weapons, bows, daggers having reduced drop rates etc., all of which effectively mirror dynamics that also existed in Personal loot.
Right now, we don’t have the exact stats for it, but it’s extremely likely that Necks/Rings/Cloaks have a significantly higher item weight in Need/Greed compared to armour of an armour type.
But any way, this is straying away from the point. Loosening trade restrictions, implementing preferences for certain players for certain items, etc. are all things that aren’t inherently tied to Need/Greed. It’s just that they’ve now been applied to Need/Greed to try and make it better. They’re improvements that could be applied to Personal loot as well. When you’re comparing Need/Greed to Personal loot, lumping those into one system or the other is skewing your comparison by including elements that aren’t inherently part of the loot system.
That said, I should point out here that you’re making a massive assumption on how Personal loot actually worked behind the scenes, which you’d need to peek into Blizzard’s code to confirm.
It’s entirely plausible that the way Personal loot actually worked was to first roll 5 items (with a filtered loot pool), then allocate them in sequence, discarding items that are invalid to be allocated to then roll a new item again, etc., until it ended up generating 5 valid items with 5 valid allocations. There are even technical reasons to do it that way.
You can say that, but what data we have on how PL works is predicted remarkably accurately by this model.
1 Like
There are many other models that would achieve the same results, or similar results, so trying to reverse-engineer something based on stats is a terrible idea. Hell, we didn’t even HAVE useful stats for Personal loot, especially if you start drilling down to specific items. Even at the end of Shadowlands, Wowhead’s recording of loot drops (as an example) had nowhere near the amounts required for statistical analysis. Many items just had nothing.
So the idea that you have whatever data, and that whatever data is somehow predicting anything accurately, or even that any prediction actually says ANYTHING about how the system actually worked, are all massive points of failure in trying to assert anything. You boldly throw out the line “what data we have on how PL works”, but you neglected to actually reference what this data even IS.
Uhh, I’d say that ~21k records is enough for analysis to be statistically significant (Painsmith, all difficulties added up)
Just curious but how do you intend to address the issue with people running lower tier content for transmogs? An easy solution would be to have all lower level transmogs unlocked for that slot when you win a tier set. So if you win a mythic piece, it automatically unlocks LFR, normal and heroic version as well. This way you remove the need for people to farm content below their current gear level just to unlock the various raid transmogs.
1 Like
That’s not how statistical significance works.
But setting that aside, if you’re pulling 21k by just adding Wowhead’s “out of” on the number of times the item was looted by someone running the Wowhead client, there’s some pretty major flaws*. Hell, it doesn’t even accurately record the “drop rate” of Ancient Anima Vessel, and that’s the one item we know should be reliable. There also obviously weren’t 162 total items (adding the number of non-Conduit items all together) that ever dropped in 683 Mythic kills. What happened to the other ~2500 items?
In what possible world are you looking at that Wowhead page and going “This seems remarkably accurate” rather than “This is utterly meaningless”?
*WHich’d require you to drill into how the Wowhead client works. How does it deal with 2 people running it in the same run? Items that are not looted inside the raid but people hearth out/join another queue and get it in their mail (fairly typical behaviour in LFR)? etc.
This comes down to how your group handles loot. If you’re getting shafted an entire tier, it’s because you’re playing with groups that don’t care about you. Under personal loot, you weren’t guaranteed anything.
In Shadowlands, I can’t count the number of nights I went to raid and got nothing but Anima. My loot came from guildmates trading me what they didn’t need because they had better. Under the new system, my guild still has the option to trade me loot. We work as a team.
In my opinion, the only people who are in a situation where it feels everyone is out to screw them are the folks who pug their way through the game. Or those who can’t trust their own team. The solution is the same: find friends. This is a massively multiplayer game where social interactions are intended. Utilize that instead of complaining that everyone is shafting you and asking for a design where other people don’t matter.
It’s going off the idea that the wowhead looter is a random sampling of the total population. When you hear about various approval ratings in the news, do you think they surveyed everyone? No, they surveyed a random representative sample.
Also, there’s a blue post that suggests that this model is correct.
Stuck waiting over an hour in a Solo Shuffle que. Still not fixed.
Day 2 of not being able to carry people through LFR to farm transmogs.
You don’t know jack about whether the Wowhead looter is even representative, and even a cursory glance at the examples that are clearly and obviously wrong results should make you throw all of it out immediately.
If you’re doing random representative sampling, you need random sampling to actually be accurate and representative. I actually did a temp job back in 2006 where I worked for companies that would do this sort of surveying, and for them to have anything meaningful, they had very specific requirements like “We need 100 responses from people aged 20-30, we need 250 responded from people aged 30-40, we also need 50 responses from someone of Maori or PI ethnicity, we need 60 responses from someone of Indian ethnicity, we need 250 responses from people in this income bracket” etc. At the end, you’d be turning down responses from interested parties to find that last person who was 60+, X ethnicity, of this height, and this weight, because that’s what you needed. You wouldn’t just throw in every result you got, because it would throw off whether your sampling was representative. Those requirements would be something people would have to compute first before they even started sampling.
THAT is the kind of work that goes into what you see in the news.
If you’re referencing the Blue post in the beta thread (either #23 or #363 in this), it did not confirm or suggest anything about “this model”, or any specific “model”. But feel free to dig up the quote/post.