|
Post by manuka on Feb 4, 2017 9:35:09 GMT
Would like any info on "random XR" and how random they really are. Are all items put into separate pools, I.e cloaks is one and motes are another. Then a random item pool is selected then a random item from that pool.
Or are all items put into one pool and one item is randomly selected.
Lasty do some items have a higher chance of being chosen, I have 3 of the same xr cloaks. With there only being 11 xr cloaks the chance of getting one should be low getting 3 of the same cloaks doesn't feel very random. I have also seen another player get this cloak, tapestry of night.
|
|
|
Post by Raj on Feb 4, 2017 11:08:47 GMT
There's been some discussion hereFunky then posted the current scriptI tried to translate the xr formula to readable numbers and my rough results have been: Out of 101 possible results (assuming motes are refund, and hopefully soon removed from the pool so don't have to ask sg help whole time ), the chance (could use a % been so close) for a specific base item to be selected are: racebook 2 weapon 10 (all types) helm 7 armor 15 (all types) boots 7 bracers 7 (gloves+gauntlets) amulet 7 belt 7 cloak 7 ring 10 shield 10 (all sizes) torch 2 And once base item has been selected, pick a specific copy from the xr list. There're some more xr items in the making but atm the wiki should cover them all (I didn't witness anything, except weapons and books, that is not on that list); chances are the same for each one, if you get multiple meh copies it's just rng hating you as usual.
|
|
|
Random XR
Feb 4, 2017 12:25:02 GMT
via mobile
Post by manuka on Feb 4, 2017 12:25:02 GMT
Cheers for the info that clears a few things up
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 5, 2017 21:09:26 GMT
I think the XR system either suffers from excessively 'sticky' RNG or there is some glitch present - not sure how many random XRs have been rolled server-wide but I believe it's still under 100 (perhaps closer to 75?)... and yet there have been 4x Tapestry of Night (Cloak), 3x Perfect Emptiness (Belt) and around 10x Motes.
In theory the chance of getting a specific Cloak or Belt should be something like 0.5-1% each, and I believe Motes were intended to be at 2-3%; to me these numbers suggest that something might be going wrong.
|
|
|
Post by condude on Feb 5, 2017 21:19:45 GMT
There's actually a formula that explains why you'd get a seemingly unexpected result like that. It's why you only need 23 people to, on balance and probability, get 2 people with the same birthday. Since there are so many items (101 IIRC), then the probability of drawing a specific item repeatedly is low, but the probability of drawing an item repeatedly is rather high. Remember that you're only looking at the outliers, when there are somewhere in the vicinity of 100 items that are rolling at a reasonably average rate.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 6, 2017 8:15:01 GMT
I'm aware of the Birthday Paradox aspect of probability, and its unintuitive nature is something that makes evaluating RNG very difficult from a player's point of view; for example, at what point of unexpected outcomes should we begin to wonder if something is actually wrong? It's certainly true that with a large enough sample size it's likely that a couple of extremely unlikely events will occur. The thing is, so far the sample size for XR rolling is relatively small (again, not sure what the server-wide total is but I would estimate it's probably 75 or less?), and even from my personal experience there were already a lot of 'coincidences'. From memory [Edit: Checked my original list, updated accordingly] Last year I made exactly 24 random XR rolls and in those I received: - 4/24 motes; ~17% frequency of an item type that should be ~2% frequency. Two of them were in a row [same reset - funny story actually, since I had just gotten it refunded by SG and then promptly needed a second refund]. - 7/24 rings; ~29% frequency of an item type that should be ~11% frequency. Three of them were in a row [same reset]. - Two of the *same* large shield ( Creation and Composition) in a row [ can't remember if it was same reset or not different resets]; the chance of getting that specific item is under 0.5%. I doubt there's anything wrong with the loot script (maybe for the Motes?), but between my own rolls and stories like the OP's it's difficult to NOT feel that the RNG behaves 'stickily' here. The unfortunate thing is that if this is true, there probably isn't a realistic/easy fix for it.
|
|
|
Post by drunkenboastor on Feb 6, 2017 18:59:01 GMT
I have spent a lot of time rerandomizing items in the forge and often see patterns happen. I am not surprised to see repetitive results for turn ins for canopics.
On 3 occasions, I have had both of the ur chests at the end of thids drop the Dragonblooded subrace. Gremlins in the random number generators?
|
|
|
Post by manuka on Feb 7, 2017 1:32:48 GMT
Out of interest poli did you buy your xrs on the same toon
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Feb 7, 2017 8:53:50 GMT
Out of interest poli did you buy your xrs on the same toon Yes, always bought on my main account toon (Druid).
|
|