@greywynd I think we disagree here fundamentally. And that's all fine. Your opinion is it currently works. To many of us it currently doesn't. I think its probably time to close this thread as the OP has made their point, some have supported it some haven't. Its time to get back in game and have fun
0
greywyndMember, NW M9 PlaytestPosts: 7,092Arc User
Not at all. You want 100% that is what the coal ward is for. If you want to gamble, there are the pres wards. Any changes to the system would simply be for easier and/or cheaper. Whether or not players actually like the system in place is a totally different animal.
I'm not looking for forgiveness, and I'm way past asking permission. Earth just lost her best defender, so we're here to fight. And if you want to stand in our way, we'll fight you too.
1
plasticbatMember, NW M9 PlaytestPosts: 12,208Arc User
edited September 2017
IMO, the current system works as this thread ask for. The problem is the price of c-ward. If the price of c-ward is (say) the price of 70 p-ward, people probably will be okay with it.
Another thing that is not working properly is RNG itself. I believe the program cache the result of RNG and that is not limited to refinement RNG. How many times have you experienced failing to open 75% professional node 4 to 5 times in a row? For me, too many times. If that 4 to 5 times are based on the same re-used random number, that would make sense.
Based on my "myth" above as a "belief" to deal with refinement/opening bag, etc, I would say I yield better than average result.
*** The game can read your mind. If you want it, you won't get it. If you don't expect to get it, you will. ***
I'm all for the suggestion. Except I think correct forumal should be ceiling(1/p)-1rather than ceiling(1/p) because one ward is not consumed. On average Cryptic will get the same amount of money via Zen. But I'm ok for ceiling(1/p), one extra ward is small cost for predictability. It is just player experience will be imporved.
Basically RNG is very bad in the game. I do not know how they implemented it, but they have done it wrong. Everyone possibly noticed that resource nodes are either open ok or fail several times. This would not happed with correct RNG implementation. Chance of second failuer is 1/16 and third one 1/64. But we are regularly see statistically improbable miracles (assuming correct RNG).
For correct implemenation I've would have stored good RNG generator in thread local variable, so it would have been fast and fair. Currently it looks like they either reseed generator often or cache results.
For refinement, I've been getting statistically better results, if I feed something to artifact gear between attempts (one white perl or green peice helps). It somewhat advance RNG. At least cases of 100 attempts for 10% stopped.
Another point that people keep forgetting is that the RNG (die) is being shared by 100's/1000's of people in succession. And we don't get to see their results. So, almost ANY attempt of personal validation for the RNG system will NOT be accurate. Only the people in the Dev/Operations area could get a look.
Another point that people keep forgetting is that the RNG (die) is being shared by 100's/1000's of people in succession. And we don't get to see their results. So, almost ANY attempt of personal validation for the RNG system will NOT be accurate. Only the people in the Dev/Operations area could get a look.
This is the problem with trying to determine whether the RNG is working properly. Anything you do fairly often you will have runs of simply bad luck. You can say the RNG isn't working but what happened before matters not to the next roll. You have to have a very large set of data.
We don't know how its generated. If we knew that then we could generate a large dataset(millions of rolls) and plot the distribution to determine if it indeed is working properly.
2
santralafaxMember, NW M9 PlaytestPosts: 2,896Arc User
edited September 2017
I would like to put a stack of pres wards into the refinement process and have it run until match. This, rather than clicking for many minutes cursing under my breath.
So with a stack of 30 pres wards, and it takes 13 to complete the task, 13 pres wards disappear. If I run out of pres wards, it should stop with the reagents intact. I can then get more pres wards or wait for another day.
I've said this elsewhere but happy to repeat here: Imho the problem with the existing pres ward system is that you fail without making any progress. A system where each failure adds 1% to your next chance of success would do away with this is and is used in other games. Introducing this now at the same time as the refinement changes would make a lot of us happier with the new ranks being introduced. Crucially I think it will improve everyone's experience including that of new players because you don't have that empty feeling of having sunk resouces into something and gotten nowhere.
* Generally it is not possible as a normal player to verify that the RNG is working or not. The sample size is just too small.
* Also remember that 1/p is the _average_ number of tries. That means you will go above this in roughly 50% of the cases, and you could actually have an infinite number of tries with no success(very low probabiity though)
* An mmorpg is calling the RNG at a furious rate. It is very unlikely each player got his own RNG state for any reason. So seen from the players point of view the RNG actually will be truly random as opposed to pseudo-random since you get a uncontrollable number of other calls to the RNG between each of your calls. The only thing that could mess up the randomness is if the values returned from the RNG are not uniform, but verifying uniformness in an RNG is fairly easy to do.
* Also remember that it is a basic part of human psychology to remember the very good or very bad things, all the average stuff in the middle is quickly forgotten. So when people swear the RNG is broken because of this or that long run, it is just their mind playing tricks and forgetting all the average tries in between
You know, the problem is that if there is label 5% chance, on average you should try 20x to refine it. For 10% it is 10, 25% 4 and so on. I usually do need double the amount of wards but I do not get any first try wins, so I guess this simple statistics does not apply here. My score this double event is 5 wards on 50% chance, 50 wards together on 2x10% chance and 35 wards on 5% chance. To sum it up: 90 wards total versus 41 according to statistics. Yeah, I was just unlucky, but as this happens every time...
I was keeping track of the number of wards Ive been using for upgrades, and the sample size wasn't awesome (maybe used 150-200 wards over 5-10 enchants) but it was almost spot on what the averages indicate it should be. On some 10% chances I went through 30-40 wards, but on others it was successful first time etc. and overall it balanced out
This double refinement weekend I upgrades two artifacts to mythic and used 21 Pres wards in total for all stages. For the last stage 5% chance, I used 8 wards for both artifacts (the second using none because it was successful first go). I'd say for every really bad result is there is likely a good one that ppl don't come to the forums to gloat about.
Those Pres wards I got from opening coffers - I opened around 130 coffers and got 5 Coal wards and 28 Pres Wards. I have found it fairly reliable that you get around one coal ward per 30 coffers (roughly).
I've been tracking this for years and over tens of thousands of attempts I'm using 20-30% more wards than I should, today 2x10%s took me 63 wards split roughly equally although earlier in 2xRP I did a 5% with 5 wards. This weekend seems particularly bad for RNG as I also took 8 attempts at 75% to mastercraft 2 items which cost me 3-500K/attempt depending on how you calculate it, almost bankrupting me as the total worth of the items was 2.3M.
I've seen something similar though I don't have the length of time in it that you have. However, I was curious to figure out why Cryptic claims the success rate they do. So I decided to track refining success rates both with and without wards. My data suggests that using a ward lowers your chance of success by 10-30%. Take the 40% chance of upgrading Rank 5 to Rank 6. To test this I refined 200 Rank 5 Quartermasters from the Seige event. The first 100 I refined without wards and had a success rate of 40.8%. Then I did the next 100 with wards and it was 31.5%. There seems to be a problem with the code that checks success if you use a ward. I don't know if it's intentional, but it's systemic, and measurable. I have seen similar rate problems at all ranks up to Rank 8, I just don't have the stacks of Greater Marks to burn on testing the rate at higher levels both with and without wards. But the success rate with wards is below the stated rates in game, that much I have seen.
2
plasticbatMember, NW M9 PlaytestPosts: 12,208Arc User
You know, the problem is that if there is label 5% chance, on average you should try 20x to refine it. For 10% it is 10, 25% 4 and so on. I usually do need double the amount of wards but I do not get any first try wins, so I guess this simple statistics does not apply here. My score this double event is 5 wards on 50% chance, 50 wards together on 2x10% chance and 35 wards on 5% chance. To sum it up: 90 wards total versus 41 according to statistics. Yeah, I was just unlucky, but as this happens every time...
I was keeping track of the number of wards Ive been using for upgrades, and the sample size wasn't awesome (maybe used 150-200 wards over 5-10 enchants) but it was almost spot on what the averages indicate it should be. On some 10% chances I went through 30-40 wards, but on others it was successful first time etc. and overall it balanced out
This double refinement weekend I upgrades two artifacts to mythic and used 21 Pres wards in total for all stages. For the last stage 5% chance, I used 8 wards for both artifacts (the second using none because it was successful first go). I'd say for every really bad result is there is likely a good one that ppl don't come to the forums to gloat about.
Those Pres wards I got from opening coffers - I opened around 130 coffers and got 5 Coal wards and 28 Pres Wards. I have found it fairly reliable that you get around one coal ward per 30 coffers (roughly).
I've been tracking this for years and over tens of thousands of attempts I'm using 20-30% more wards than I should, today 2x10%s took me 63 wards split roughly equally although earlier in 2xRP I did a 5% with 5 wards. This weekend seems particularly bad for RNG as I also took 8 attempts at 75% to mastercraft 2 items which cost me 3-500K/attempt depending on how you calculate it, almost bankrupting me as the total worth of the items was 2.3M.
I've seen something similar though I don't have the length of time in it that you have. However, I was curious to figure out why Cryptic claims the success rate they do. So I decided to track refining success rates both with and without wards. My data suggests that using a ward lowers your chance of success by 10-30%. Take the 40% chance of upgrading Rank 5 to Rank 6. To test this I refined 200 Rank 5 Quartermasters from the Seige event. The first 100 I refined without wards and had a success rate of 40.8%. Then I did the next 100 with wards and it was 31.5%. There seems to be a problem with the code that checks success if you use a ward. I don't know if it's intentional, but it's systemic, and measurable. I have seen similar rate problems at all ranks up to Rank 8, I just don't have the stacks of Greater Marks to burn on testing the rate at higher levels both with and without wards. But the success rate with wards is below the stated rates in game, that much I have seen.
If you really want to see what happen without burning your marks, go to the test server to check that out.
*** The game can read your mind. If you want it, you won't get it. If you don't expect to get it, you will. ***
On this last double refinement we had I took 11 enchants from 11 to 12 and 4 artifacts from Legendary to Mythic with 500 pres wards. Figured I was on a hot streak the next enchant that I tried took 216 pres wards to go from 11 to 12. In the end I guess it all evens out. On a different note tho, I have 9 alts max pray almost every day and I have not got a single Coal Ward from a box since I came back to the game last Christmas, 10 months...... thats some craptastic odds. Good thing I didn't have to lvl up my weapon enchant.
I'm skipping most of the posts, so forgive me if I missed something, but it looks like the original post was asking for a system similar to Vorici, Master Assassin from Path of Exile. This isn't cheap and guaranteed. It is expensive for the given tier of item desired, but also proportional to the chance.
A c-ward is considered expensive for any upgrade with a chance above like 3%, however, it is also basically throwing money away to use a c-ward on a 50% chance. What it sounds like is what Vorici allows in Path of Exile. He will take incredibly more resources than the average to guarantee the results. An example of this using Neverwinter chances and wards would be:
These are all pretty crazy expensive compared to the average, but relate better to the chance of success than a c-ward would, since c-wards have a fairly static value that hovers around 90 pres wards.
Did Vorici/Path of Exile have another means to guarantee success besides the massive resource route?
You could use the resources naturally to RNG it. The same resources you would have overspent Vorici to guarantee the results. In this case it would be using the p-wards naturally.
Comments
Another thing that is not working properly is RNG itself. I believe the program cache the result of RNG and that is not limited to refinement RNG. How many times have you experienced failing to open 75% professional node 4 to 5 times in a row? For me, too many times. If that 4 to 5 times are based on the same re-used random number, that would make sense.
Based on my "myth" above as a "belief" to deal with refinement/opening bag, etc, I would say I yield better than average result.
Basically RNG is very bad in the game. I do not know how they implemented it, but they have done it wrong. Everyone possibly noticed that resource nodes are either open ok or fail several times. This would not happed with correct RNG implementation. Chance of second failuer is 1/16 and third one 1/64. But we are regularly see statistically improbable miracles (assuming correct RNG).
For correct implemenation I've would have stored good RNG generator in thread local variable, so it would have been fast and fair. Currently it looks like they either reseed generator often or cache results.
For refinement, I've been getting statistically better results, if I feed something to artifact gear between attempts (one white perl or green peice helps). It somewhat advance RNG. At least cases of 100 attempts for 10% stopped.
We don't know how its generated. If we knew that then we could generate a large dataset(millions of rolls) and plot the distribution to determine if it indeed is working properly.
So with a stack of 30 pres wards, and it takes 13 to complete the task, 13 pres wards disappear. If I run out of pres wards, it should stop with the reagents intact. I can then get more pres wards or wait for another day.
Imho the problem with the existing pres ward system is that you fail without making any progress.
A system where each failure adds 1% to your next chance of success would do away with this is and is used in other games.
Introducing this now at the same time as the refinement changes would make a lot of us happier with the new ranks being introduced. Crucially I think it will improve everyone's experience including that of new players because you don't have that empty feeling of having sunk resouces into something and gotten nowhere.
* Also remember that 1/p is the _average_ number of tries. That means you will go above this in roughly 50% of the cases,
and you could actually have an infinite number of tries with no success(very low probabiity though)
* An mmorpg is calling the RNG at a furious rate. It is very unlikely each player got his own RNG state for any reason. So seen from the players point of view the RNG actually will be truly random as opposed to pseudo-random since you get a uncontrollable number of other calls to the RNG between each of your calls. The only thing that could mess up the randomness is if the values returned from the RNG are not uniform, but verifying uniformness in an RNG is fairly easy to do.
* Also remember that it is a basic part of human psychology to remember the very good or very bad things, all the average stuff in the middle is quickly forgotten. So when people swear the RNG is broken because of this or that long run, it is just their mind playing tricks and forgetting all the average tries in between
A c-ward is considered expensive for any upgrade with a chance above like 3%, however, it is also basically throwing money away to use a c-ward on a 50% chance. What it sounds like is what Vorici allows in Path of Exile. He will take incredibly more resources than the average to guarantee the results. An example of this using Neverwinter chances and wards would be:
90% - 3 wards
50% - 8 wards
25% - 16 wards
10% - 40 wards
5% - 90 wards
1% - 140 wards
These are all pretty crazy expensive compared to the average, but relate better to the chance of success than a c-ward would, since c-wards have a fairly static value that hovers around 90 pres wards.
Signature [WIP] - tyvm John
Signature [WIP] - tyvm John