Serious problems with Elemental Armor..sigh

[Renjah]

I am wondering is the %mods have any effect when used on PoP combines. I have noticed MUCH better success rates on pottery from when I was 240 and 241 skill (both 252 with trophy) and I get more successes with 250 skill (252 with trophy).

[Silound]

Gurck, your example of a coin won't quite fit for this, but you have the right idea.

Anything that gives a probability of success/failure, 0/1, true/false, yes/no, or any other permutation of options isn't going to achieve randomness, simply because there are only 2 options, and if it isn't one, it's the other.

My memory is fuzzy past two, been a long time since I took Theoretical Design, but the gist is anything that does not have an infinite number of possibilities is therefore a probability, and by law of probability, eventually, there will be an even number of each option.

Thus the RNG streaks because it's really a probability equation, otherwise known to programmers and design specialists as a pseudo-random number generator.

[Sunfire]

It is well known that EQ has about the least random RNG in existence - it was actually "TWEAKED" once about a year into the game (before which it was using the most basic Random function you can imagine). From what I understand, EQII's is a completely different function from a completely different source if that tells you anything.

That being said - I also believe some skills, smithing being one, are operating at a higher cap now than 252 - something like the 350 which was mentioned.

Practical advice: If you fail ONCE - do 20 or so easy combines to move the RNG's seed number then go back to the good stuff. I have been doing this for my PoP combines and I have yet to get back to back failures whereas I had often experienced them before.

And by the way - you get a big /comfort for all the BPs

[splunge]

/em tosses another post in the RNG bin

I can understant why youre upset. Failing 10/11 very very expensive combines is rough.

Still, in every possible subset of a RANDOM set, streaks like this are an absolute guarentee to happen at some point. It may not occur as often as, say 5 or 6 out of 11, bt it happens nonetheless.

Its also human nature to bemoan tragedy far far greater then rejoyce triumph. For every pwerson who went 1/11, theres 5 people who went 10/11 and didnt post anything anywhere.

*shrug*

I hope you are fare more successful in your future combines. =)

[Nolrog]

I had a teacher in school one time (probability course) that told us to go home and flip a coin 100 times and write down the outcomes. Based on what the students turned in, he could tell who actually did the assignment, and who just made it up. The ones that did the assignment had streaks of 10 or 12 heads (or tails) in there somewhere. The ones who made it up had streaks of no more than 3 or 4. As someone mentioned, you can expect to have streakes like this of failures.

I also remember reading somewhere that your probability of success is not at it's maximum when you hit the trivial level. You need to be quite a bit higher. So if you are making something that trivials over 250 (and no one knows the exact trivial level) then it stands to reason that your probability for success would be less than expected, considering everything I've read says your skill caps out at 252.

[Mainiac]

/petiton Please stir the entropy pool for the RNG. Thank you.

[xaanru]

Well, this isn't an elemental plate example, but demonstrates the streakiness of the RNG (which is normal, imo). I aggreed to make some heraldic for a guildmate (free of course). He brought me enough for several combines on each piece he wanted. I wend 6 of 9 on his armor. He was so happy he said i could keep the extra.

Between what he had left, and some scraps in my bank, I had enough stuff for 6 more combines (woot profit time). Went ZERO for 6.

Well the overall average 6/15 is right where it should be. But it didn't feel like it, as I failed the six pieces I was hoping to sell. Glad I managed to gear out my guildmate though

[Cavel]

Xaanru,

Happens to me all the time... Bring me enough to make 2 attempts at Item X... Make all of item X that the person bringing me the stuff wants with few failures THEN attempt to make some for myself and NO DICE... I thought I was the only one LOL.

[Aniien]

As I was told

At the Fan Faire I was told by several developers that the Planes of Power trade skills above 250 which includes many, have a build-in failure, none of the developers would elaborate on the percentage of the failure

A friend who smiths, tailors, fletches and bakes for a very large elemental guild, believes it is about 50% for him.

[Drakkiinn]

fun with the RNG

a fun way to shove the designers nose in the mud...
Two PC's with identical stats (generaly mules made for the purpose)
train 1 to 100 in brewing leave the other at 0

Both attempt 100 Combines of mino hero brew... I got 18 with skilled mule and 15 with unskilled sucsessful. giving me about 15-20% sucsess 100+pnts out of trivial But was VERY rudementery and small test.
the RNG sucks VI knows it but doesn't care, it's not their 10kpp after all.

BTW expect lots of failures on something more than 100pnts above your trivial as with elemental plate.

[kiztent]

Originally posted by Nolrog
I also remember reading somewhere that your probability of success is not at it's maximum when you hit the trivial level. You need to be quite a bit higher. So if you are making something that trivials over 250 (and no one knows the exact trivial level) then it stands to reason that your probability for success would be less than expected, considering everything I've read says your skill caps out at 252.

No Nolrog, that's not true. It's true for items with low trivial, but for items with trivials over 180 (somewhere in that area) you get the max success rate BEFORE the item trivials.

Under the old success rate anyway.

[Viacios]

Working as intended

Duh! The items you are attempting were designed to be combined by a GRANDMASTER, aka actual 250 skill with a skill bonus item. Hence the 250+ trivial. In our guild only someone with 250 skill and smithing item is allowed to have the components for a combine. And the success rate is about 50%+ Planer armor was not intended to combine easily, and especially not by non grandmasters.

[TheGiantOne]

Okay...

I have made Farwater and all below it for elemental planes armors and overall have hit a 40% ratio of success.

But at times its 2 out of 7 and next time is 3 out of 3. You never know. My rule of thumb is if you fail the first 2 combines stop go get some coffee or a snack and try later.

But its all about the RNG remember. It can be nice and give you 100% or just outright mean (been there and done that) and give you 0%.

Also remember those elemental combines are way above 300 also. Since the new UI have not tested how much they are but I would imagine 350+ on a trivial scale.

[Nebaccanezzar]

Im in absolute shock no programmers of any kind have posted on this thing yet. Here is the simple fact that is frustrating all of you so much. Computers can NEVER EVER create a truly random number. All random number generators must follow an algorythm, which in itself is a pattern, and there for can never be truly random. The everquest RNG is even worse than most IMO, I belive they wrote the code for it themselves, even the random library in old ANSI C is better hehe. This is just one of those things that we just gotta suck up and deal with.

Nefandus Servito'Acrebus
65 Arch Lich Druzzil Ro
(Great Adventure Partizan)
Tailoring 205, Baking 240,
Fletching 220, Jewcraft 235,
Brewing 240, Pottery 223,
Smithing 222, Tinkering 200

[Laeelin]

Random talks always remind me of the Monty Hall problem =)

Most people dont understand what real random numbers are (or probility)


This is the Problem:
=================

You are on a game show. You are presented with a choice of 3 doors: behind one is a luxury car, and behind the other two are nothing. The host asks you pick one of the doors.

After you do this, as part of the game he opens one unpicked doors which he knows is empty. There are now only the door you picked and one remaining door which are unopened. You are asked if you would like to switch your choice. Should you switch?

==================

Almost everyone doesnt thing that changeing or not changeing makes any real difrence...

But it does.

It makes a HUGE diffrence.

if you change you have a 2/3 chance of winning
if you dont change you have a 1/3 chance of winning =)

Normally the only way to really prove it is to actually set up a situation like that and do 25 or so runs to prove that changeing helps.



I guess the poing of this is that random isnt what most people think it is any more than probility is what most people think it is.