test content
Logo
What is the Arc Client?
Install Arc
Take the RNG out of mastercrafting
feanor70118
Member, NW M9 Playtest Posts: 1,170 Arc User
Being on my tenth attempt to make a copper coin, it seems apparent that having this system could and should be much more sensible than gambling on a 35% chance of success even with a mythic Forgehammer of Gond and ahving to go earn and spend 2100 guidl marks every time an attempt fails. Over and over and over and over and over and over and over and over and over.
Instead of having a quest to make one item with a maximum (unless one has the unobtainable adamantine profession tools - which by the way would be able to be crafted from mithral ones if professions made any sense) 35% success rate which, in Cryptic RNG terms, is often less than 10% in actuality, have a quest to make 105 items. The 35% chance now would translate into a guaranteed 35 items. Without a mythic forgehammer one would be able to make 20 items at a time. But, everyone would have a clear and certain path to finishing mastercrafting instead of endlessly wrestling with the RNG.
I have no idea why Cryptic seems to think that so many questions in this game should be answered by "frustrate players with the RNG," but it's the wrong answer.
Instead of having a quest to make one item with a maximum (unless one has the unobtainable adamantine profession tools - which by the way would be able to be crafted from mithral ones if professions made any sense) 35% success rate which, in Cryptic RNG terms, is often less than 10% in actuality, have a quest to make 105 items. The 35% chance now would translate into a guaranteed 35 items. Without a mythic forgehammer one would be able to make 20 items at a time. But, everyone would have a clear and certain path to finishing mastercrafting instead of endlessly wrestling with the RNG.
I have no idea why Cryptic seems to think that so many questions in this game should be answered by "frustrate players with the RNG," but it's the wrong answer.
5
Comments
But I agree that the current % success / fail system is frustrating for players who get a streak of bad luck. Some kind of mechanic to ensure progress towards an eventual success with each failure would be nice to offset player frustration with streaks of bad luck... but it would also necessarily devalue the eventual products (if stuff is easier to make / fewer resources spend on average, then the prices would go down).
It's also true that the 35% success rate is really punitive for unlocking MWs. But that's part of the point, again. It keeps MWs less accessible due to the risk / cost involved, which again keeps supply a bit shorter and prices a bit higher.
Every change that makes MWs more accessible for unlocking / crafting would necessarily also decrease the profitability for the mastercrafters.
In general, there are times when I get streaks of bad luck (having do do 50+ p. wards on a 10% upgrade chance). There are times when I get lucky (1st try on a 5% upgrade chance). I've even seen someone get a fluke wep enchant out of 4 shards that they refined without a ward.
When I did my own tests on artifacts, and asked my guildies to do the same for 2x RP about a year ago, our results were generally in line with expectations. Some folks were luckier, others were unluckier, but in the grand scheme the success rates were not so far off as to suggest the posted % are incorrect to a statistically significant degree.
Years ago I read about how RNG is implemented in online games, sadly I didn't bookmark that article and was not able to find it again eversince. What it said in general lines (personaly I don't know much of statistics and chances theory so forgive me for mistakes) was two main approaches on the issue. Each one had a distinct name which I sadly can't remember.
Assuming a result has certain chances to happen (for example 5%). When you triger the process of RNG the game draws a number from a bucket where certain numbers correspond to fail and others to success. Let's assume that 100 numbers are inside the bucket. Numbers 1-95 result in fail while number 95-100 result in success. Let's assume that RNG picked number 25 which results in a fail. The difference between cases is in what happens next.
Case 1: The number 25 is thrown back into the bucket. This means that in every RNG trigger for said event, the chance will always be 5%, no matter how many trigers you do.
Case 2: The number 25 is removed from the bucket. This means that next RNG triger will have slightly increased chances of success and that will increase with every fail.
Case 1 is pure randomness and there are no safeguards that any descent results will happen. Fails are usually so big that would drive most players away unless the chances are somewhat big (like 20%+). Imagine what would be the real chances to see any 1% chance success even after hundreds of trigers.
Case 2 is more player friendly but has to be applied to each item separately. For example, if you have two identical items and fail refining 10 times on one of them, this item will have bigger chances to succeed from the one that was never tried. I have found this kind of system in a game I played in the past. If a player failed to enchant a piece of armor or weapon, a counter appeared on said item indicating how much boost in chances it got for the next enchant (ok, it was small but it was something). That counter was reset after a success.
Another important thing in the whole RNG business is the size of the bucket aka. how many "numbers" it contains. If a result is 10% likely to happen and the bucket has 10 numbers and only 1 is the success, it means that worst case scenario is you make 9 fails and the next try is a guaranteed success. In case where the bucket is 100 numbers and 10 numbers correspond to success, things are different. Worst case scenario is 90 fails before you get a guaranteed success, although the absolut success rate is still the same (10%) for both cases.
If I have to make a guess about the RNG in this game, I'd say that it uses Case 2 mechanics but with a large number of "numbers" in the bucket. If this is the case, Cryptic could reduce those numbers to make things smoother. For crafting purposes (like masterwork crafting) where there is no existing item but one is about to be created, a method I can think to increase the chances is this: in every failed attempt to create an item, you get a "broken" version of this item. These "broken" items can be used as a reagent in the next crafting attempt to increase the chances of success.
RNG needs to be toned down. Period.
Make an item like a pres ward, you lose it but get your mats back.
With the mad discrepancy between tasks as to what you lose on a fail, being able to guarantee you don't lose a fartouched residuum or artisan's enamel would be handy. The price discrepancy between artisan's putty and artisan's enamel puts an artificial price discrepancy between metal and non metal armors. Similarly the 3 fartouched residuum (and risk of losing one on a fail) on CW/SW weapons makes them ludicrously more expensive than other classes. TRs and HRs get the cheapish end of both.
Want to make a thingamebob?
You have a 35% chance with a Gond - make this a necessary ingredient and triple the time & ingredients required.
No Gond and purples make those purples necessary and have 5x the ingredients and it takes 5 times as long.
Make it single task only.
This should make no difference to the economy in the long run as with a big sample size this is the time & ingredients required, but it would remove the frustration, nobody remembers when they got lucky, just the times they weren't.
What Class Are You?
I have a stats degree, and have done a fair amount of testing on the RNG, and over tens of thousands of tries, I consistently find the success rate for stuff is a good 20-30% lower than it should be. My research suggests that the probabilty of a failure after a success is not the same as the probability of a failure after a failure if the trials are done close together. Also there are circumstances where it appears that the same RNG seed is reused.
The sort of stupidity I'm noticing is that (because I have a load of char bound blue marks of potency), I'm taking a load of enchs 5-7, and not a single 30 or 40% chance works first time out of 20. This has happened several times.
Recent attempts at MC2 crafting I got 39/70 at 75%, this was the worst run I've had, and spent 2 days trying and failing to make a single conc aqua regia, one of the two crafts always failed. My MC3 record is pretty much spot on 75% on 250 attempts my overall MC2 record is 258/370 which is quite a bit down on where it should be on that many tries.