I've created about 20 rank 10 enchantments now, all of them were initially created from rank 5's so needless to say i have a wealth of sample size under my feet. there is absolutely, positively no question that the RNG is messed up in this game. Ive learned how to make the most of it, but OP is definitely not alone and until you learn how to combine "properly" in this game, youre gonna keep failing 20+ times in a row.
I've created about 20 rank 10 enchantments now, all of them were initially created from rank 5's so needless to say i have a wealth of sample size under my feet. there is absolutely, positively no question that the RNG is messed up in this game. Ive learned how to make the most of it, but OP is definitely not alone and until you learn how to combine "properly" in this game, youre gonna keep failing 20+ times in a row.
I am definately not wrong. whether you succeed or fail on one attempt is irrelevant to the next attempt. Every time the OP goes to upgrade the enchantment there is a 30% chance of success, this will not change no matter how many times you attempt the upgrade.
Lets say for example this RNG uses the numbers 1-100 as its range. lets set success to 1-30 and failure from 31-100. If I roll 35 and 97 on the first and second attempts there is no gaurantee that my next roll will be 30 or below. That roll has 100 possible outcomes with only 30 of those being in the "success range". The other two attempts have no factor in this.
Sure the probability chance of only getting 2 out of 26 is way lower than the statistical mean but theoretically you could make 100 attempts and still not get any successes.
Okay how you figure the cumulative chance of a success of INDEPENDENT rolls is to use the inverse of the probability and it is exponential. So if the chance to succeed is 3 in 10 or .3 we take that and subtract is from 1 and get .7. so now we calculate the chance of NOT succeeding, and we see that x = 1-(3/10)^n where n is the number of tries and x is the probability of failing. with this formula we find that at 30% and 6 tries the chance to succeed is 88%. i.e 1-(.7)^6, with 8 tries it becomes 95% and so forth. so with 24 tries and successes we find that 24/3 = 8 so basically every 8th try you got a success, and that fits the probability calculation here. You need to remember that at any given time something like 10k players are using the RNG calculator, this gives a large enough sample size that the law of averages keeps the RNG pretty even.
no that doesn't work. the formula you stated will never equal 100%, so there can never ever be a gauranteed success. There will always be that chance that you do not get the success. If you are going for the average than yes evey 8th attempt should be a success but it is never an absolute that it will be.
every attempt you make is still the base 30% chance to be a success
As a test i went to www.random.org and performed a few tests with the following paramaters: 26 results with a range between 1-100 (inclusive):
test 1: 7 results =< 30
test 2: 8 results =< 30
test 3: 5 results =< 30
test 4: 12 results =< 30
test 5: 6 results =< 30
test 6: 10 results =< 30
test 7: 8 results =< 30
test 8: 3 results =< 30
test 9: 9 results =< 30
test 10: 8 results =< 30
As you can see the average of those tests is close to what you are saying (7.6 compared to 8) but notice the high result of 12 and low result of 3.
I think I have the best luck refining stuff, the trick is to pick the right spot to stand in, you have to be on your mount, and can't be at a campfire. I managed to fuse a rank 10 first go, and fused so many rank 7s pre-mod 2 with great success rate.
All you have to do it pay me 200 zen and I'll tell you the secrets.
(ok, that was a joke, I'm not accepting any payments just in case it's in violation of some forum rules that I'm not aware of)
Have the gear and skills, but lack the friends to play with? Come and apply for Essence of Aggression. We have been here and strong since beta. (Immature, rude, and arrogant people will not be accepted)
You have not shown anything. All you did was point out that it is not impossible for it to occur. Unless you looked through their code and saw that there were no uninitialized variables, improper seeded, etc etc you have not proven that it wasn't a bug.
Faaaaaaaaaaair enough, I agree.
To be honest, using pseudorandom number generators without a failsafe is "buggy".
Think about for a sec. Proper game design that allows a 70% chance of success should be designed around a normal sample size that any player will encounter. What I mean is, no one except bots refine 1M items. So who cares if it's a 70% success over 1M. What I do care about is how likely is it for me to get 0% success over a normal sample size of 10.
That is where a failsafe comes in. As a game designer, do you really want folks to get 0 successes out of 10 tries when something has a 70% chance for success? Yes mathematically that is possible even with a perfect true random number generator, but that is not a proper design.
That is EXACTLY the point I made (or I tried to) when talking about that strategy game. So I guess we are in the same boat after all.
What people forget also is that RNG requires an input to randomise, most use the current CPU clock count as this constantly changes, however something on the scale of NW can't use the clock count because it would bog the server down too much, so most MMOs use the current time as a base for the randomiser. (For reasons why not to use the clients CPU count the main one is that someone could alter their clock count to a 'luckier' count)
This of course leads to the fact that there are certain times of the day that are luckier than others as that time has more chance of generating the right numbers, so if you're failing a lot, stop and try again later.
0
bioshrikeMember, NW M9 PlaytestPosts: 4,729Arc User
edited February 2014
What is needed is a bit of transparency. For instance, when I use a skill kit, show my roll and show the target number that was needed. I'd even go so far as to maybe show a history of the last 10 or so random rolls that pertain to me - if a predictable pattern related to one factor or another comes to light, then it is an opportunity for improvement. Showing the numbers would also help identify more obvious problems - like if a skill node says I have a 75% chance for success, but my target number is 76 or higher on a roll of 1-100, then the chance for success is obviously mislabeled...
<::::::::::::::)xxxo <::::::::::::::)xxxo <::::::::::::)xxxxxxxx(:::::::::::> oxxx(::::::::::::::> oxxx(::::::::::::::> "Is it better to be feared or respected? I say, is it too much to ask for both?" -Tony Stark Official NW_Legit_Community Forums
Comments
RNG is wrong as in %, or as in 'streaky' or both?
Lets say for example this RNG uses the numbers 1-100 as its range. lets set success to 1-30 and failure from 31-100. If I roll 35 and 97 on the first and second attempts there is no gaurantee that my next roll will be 30 or below. That roll has 100 possible outcomes with only 30 of those being in the "success range". The other two attempts have no factor in this.
Sure the probability chance of only getting 2 out of 26 is way lower than the statistical mean but theoretically you could make 100 attempts and still not get any successes.
every attempt you make is still the base 30% chance to be a success
As a test i went to www.random.org and performed a few tests with the following paramaters: 26 results with a range between 1-100 (inclusive):
test 1: 7 results =< 30
test 2: 8 results =< 30
test 3: 5 results =< 30
test 4: 12 results =< 30
test 5: 6 results =< 30
test 6: 10 results =< 30
test 7: 8 results =< 30
test 8: 3 results =< 30
test 9: 9 results =< 30
test 10: 8 results =< 30
As you can see the average of those tests is close to what you are saying (7.6 compared to 8) but notice the high result of 12 and low result of 3.
All you have to do it pay me 200 zen and I'll tell you the secrets.
(ok, that was a joke, I'm not accepting any payments just in case it's in violation of some forum rules that I'm not aware of)
Have the gear and skills, but lack the friends to play with? Come and apply for Essence of Aggression. We have been here and strong since beta. (Immature, rude, and arrogant people will not be accepted)
This of course leads to the fact that there are certain times of the day that are luckier than others as that time has more chance of generating the right numbers, so if you're failing a lot, stop and try again later.
"Is it better to be feared or respected? I say, is it too much to ask for both?" -Tony Stark
Official NW_Legit_Community Forums