Ever since I have "improved" my suit to 930 luck I have been getting shitty loot. I was getting pretty good stuff when I used a 800 luck suit. Is it normal or there are "magical" numbers that I don't know of?

I am hearing a few people say that. I have a luck suit that is 1140 medable and I farm a fair amount. I find my counterparts who farm in the 800's get a lot of really nice stuff like 1/3 and 1/2 jewelery while i get consistently high resists pieces of armour with higher misc jewlery and luck stuff. I pretty much only farm bloods.

I have a theory but I haven't been able to find anything that supports it. Here it is though.....At my level of luck I have about a 50/50 chance of getting the bump. At 800 or so its significantly lower. My idea is that since the chance of getting the bump is lower at 800 the actual bump itself on the item is higher. That is the only thing I can guess is happening when people with 800 luck are getting valuable items all the time like 1/3 stuff while in all my farming I have never gotten one.

My other theory is I just have crappy luck no matter what my suit is

quote:

---

My other theory is I just have crappy luck no matter what my suit is

---

quote:

---

Originally posted by Pearl

quote:

---

My other theory is I just have crappy luck no matter what my suit is

---

I think this one's the right one.

Its called luck for a reason...

---

Thank's for the replies! I have switched back to the 800 luck suit and there is a clear difference in the loot I am getting. I hunt bloods for items and good stuff have been popping back up after I switch back. The other thing I notice is that I don't get the really high intensity stuff that pops up once in a very long while when I used the 900 luck suit. So I am wondering for all those who have 1000+ luck suit, it there really a big difference in loot after 1000 luck??

I did extensive research on this when AOS first came out. For two weeks solid I did nothing but test the effects of luck. Say what you want folks....but 800 luck is what works best for me. But it's like you said. It depends on what you want. That super high luck will help you get that UBER item. But 800 luck consistently gets you decent shit. Another THEROY of mine was that anything over the 100th mark seems to matter. E.g. 820 luck is better than 810. It seemed that my 860 suit was WAY worse than my 800 luck suit. Almost like it was just giving me credit for 60 luck. I dunno....we all basically don't know shit about it and it is just that....LUCK. Those were just MY theories and findings.

I have 972 luck currently and for the amount of time I camp bloods, I get crap loot. So, I started thinking of different parameters that may influence the whole loot property/intensity- luck relationship.

Let me start by saying I think OSI uses a crappy random number generator (RNG). In fact, I think Mr Tact once said it was a simple call to rand()!!! While practical/quick, rand() has issues. Ok, so think about how many calls to the RNG are required...anything that requires a "dice roll" would be a call to the RNG. So, for even a single person playing on a shard that is still alot of RNG calls.

Now, factor in hundreds/thousands of people on a shard all calling the RNG!!! The calls to RNG would be less than microseconds apart! So, whatever seed the RNG is using HAS TO BE able to simulate "instant randomness". As we all know, conventional RNGs are only considered pseudo-random generators because they are not truely statisically random number generators, ie they do not produce "instant randomness". For example, most use clock time as the seed or CPU temperature to 8 decimal places.

The problem with using sequential seeds is some "sequence clusters" of seed numbers all produce the same or roughly the same random number! Assume they are using clock time. So, the seeds sent to RNG might be something like 3245678 3245679 3245680 etc... If those seeds are ones in a "sequence cluster" and alot of calls to the RNG are being made, then the features controlled by luck would all be similiar!

So, this would be why loot is usually similiar as in multiple helms, leg armor, jewelry, weapons, etc... in the same kill. As a population grows on a shard (throughout the day/night) you will notice this occurring more frequently. This to me is an indicator that more calls are being made to the RNG and "sequence clusters" are being encountered more frequently.

This same concept can be applied to the number of properties on an item and the range of intensity on the items. Recall that luck controls 1) bump in number of items in a kill, 2) bump in number of properties per item, and 3) bump in intensity value for each property of each item!!! Damn, that is alot of RNG calls for a single kill!!!!!!!!!!!!!!!!!!

Ok, so what about lower luck getting better items? Simple. Recall that the higher the luck the more BUMPS you will get in 1)-3). So, if you have a lower luck, the bumps occur less frequently. Fewer calls to the RNG (especially when the shard is busy) will produce a longer distance between seed values and avoid the occurrence of "sequence clusters".

I really wonder if OSI has any concept this is occurring? This whole RNG problem is why OSI has so much trouble with bugs, when releasing a new publish even after extensive QA and limited live shard testing. The phenomemon I describe may require the full weight of a busy shard to see, and OSI just doesnt have that population availability, for testing, prior to publish to all live shards.

The real unknown is how OSI handles so many calls to the RNG? If they queue the calls it creates a whole additional level of confusion. If they use calls to different RNG functions depending on use it will help, but not fix the problem. For example, combat_rng(), loot_rng(), etc... Bottom line is the multiple calls are still ultimately to rand() *eewwwww*

So, what you really want to know is how to use this information to maximize your loot quality...
Hell if I know :-) but I will tell you what I do...

BASIC RULE...you want long seed sequence distances for anything that is sent to the RNG regarding your loot :-)

1) You can keep your luck high, but the higher the luck the more calls to the RNG are made each kill thus causing all loot in the kill to be similiar, BUT the intensities should be higher.

2) Play when the shard is not busy.

3) Of course monsters have set loot templates, so find the toughest monsters dropping the best loot to begin with.

So, I hope these tips can help you "create your own luck" ;-)

O, and yes I do have a few graduate degrees in areas related to what I discuss above *laugh* I have in fact run into the "sequence clusters" problem in my own research 8-( and it sucked hard!

Wait a minute, I actually sortof followed that...

Wouldnt that mean that the optimal value for luck would be between 800-900 though and not over 1000? Think about it, you want to be able to get the really great intensity stuff which starts at around 800, but you don't necessarily want to make too many RNG calls for number of properties, number of items, etc.. which definately occurs over 1000.

At 800 I notice the intensity changes for loot, at 1000 my experience (hunting with those with +1000) is that the number of items goes back up as well as the intensity but the distribution of intensity gets placed on all of them, making each item have a nice property but none of them uber. So by applying what you are saying to what we have seen occur, it seems like at around 800 you are not going to get too many items but you still get the possibility for the RNG to pull you back high intensity, thus you would increase your chances of seeing one high intensity item instead of several medium intensity items.

Because it takes more than one high intensity attribute to make a single item good, but if you spread out those high intensity attributes among more items it makes the quality of the loot less than it would be if the RNG was applying the high intensity to few objects.

Did that make sense?

Yes I think it does. Unfortunately that means I now am going to have to be really ticked at the effort I put into making my uber luck suit

I thought (remembering some crap that a dev spouted in passing when *first* describing luck) that the higher luck you have the better the item is likely to be, but the higher the luck you have the less likely you are to get more items...

This meant that there was a balancing act between going for a few really good items or lots of pretty good items.

Of couse, they probably implemented this but the "relations" team don't know, care or understand and are confusing people as usual...

[Edited on 29/8/2003 by Manuel : naff typing skills... ]

OK! I did more testing with my 800 luck and 900 luck suit. (couldn't quite get to 1000 luck yet) I found that I can get higher intensity stuff with the 900 luck suit but only once in a really full moon. However with an 800 luck suit I can pull decent stuff in about one third of the time. So I think 900 luck is kind of the beginning "phase" for better roll but the % of getting good roll doesn't improve until you get to 1000+ luck.

Since enchancing armour and weapon is also another "mystery", I am crossing my fingers until I can gather more high luck pieces to "destroy".

I am going to back track and test this too, I have 977 luck on my suit. I drop about 100 Blood eles a run and get nothing special.

I also see monsters loot tables as broken too. Ever cap a Shadow Wyrm only to find a regular ring or bracelet on it? wtf is that? Broken Loot table is what..
Balrons too have crumby loot in comparison to the lesser Blood Elementals, Ancient Wyrms too seem worthless for effort.

I found most of my killer items at about 500 luck to be honest, that is when I got good luck ranges and excellent LRC items. I havent seen a LRC above 14 in over 500 Blood Elemental Kills.

Just bringing this back up, to see if anyone else has worked on it?

THanks!

S

Hmm i remember i go my uber item with only about 600 luck
It was a 64 res Meddable Gorget..
Not sure how "uber" this is?

You know guys, this whole luck thing always pissed me off. Why the hell they invented luck, I don't know. Unfortunately, luck suits give 0 resists, and they are only for tamers... I have a luck medable suit that reaches 650 luck but my provoker cannot wear it, unless I fight ogre lords. Rest of the monsters, cast a spell on you and half your energy bar is gone Anyway, I always show up with 250 luck and 70 fire resist, doing my T-hunts that thank god are not based on luck...

Point of this post: luck sucks, camping of monsters to get some good items sucks, item based gameplay sucks!

Yea, umm. Ok.

So anyone ELSE done any testing?

S


We can complain about the game, but complaining about people playing the game the way the people that MADE the game MAKE them play the game... Nevermind, I'm confused.

http://boards.powergamers.net/viewthread.php?tid=12477 />

Manual, this is absolutely untrue. If you can find the table it would go a long way to proving OSI wrong because on their website they specifically say that LUCK influences the amount of items you can get in the loot depending on monster template!

Check out what OSI says here:
http://guide.uo.com/itemproperties_4.html />