Greetings, glorious adventurers! If you're joining in our Alpha One spot testing, please follow the steps here to see all the latest test info on our forums and Discord!

Don't kill the game with too much RNG

Some games have way too much RNG in them. For example, crafting in some games gives forces you to craft a certain item with "Random secondary stats" or "Random stats" in general. This is okay for a very limited degree. You don't wanna farm materials for weeks and months to craft something that in the end is worthless because they got very bad stats. One solution is to only randomize niche stats or special stats that are not game breaking when considering crafting. 


Also, don't have too much RNG on boss loot. A boss that drops random items and rare items is fine. But, a boss that has a chance to drop random items with random stats is just a pain in the ass. Once again this can be solved by only allowing randomness on minor secondary stats and nothing that affects your performance noticeably. 

One thing to really stay away from is complete random loot system from grinds like PvP. In World of Warcraft for example they have this system that is universally hated and costed WoW many subscribers because it is so unbelievably stupid. This is how it works: 

1.Each instanced PVP Battleground that has a 5-10 min queue and lasts for 10-20 min has a 30% chance of giving you a box.
2.This box in turn has a 10-20% chance of giving you a random item that can be anywhere from useless to good. It is also completely random which type of gearslot you end up getting (you have 12-14 gear slots). Also, there is no protection from getting duplicate gear types forever. 
3. This item in turn has random stats. If these are wrong you often destroy the item.
4. Some  items have random special abilities that are crucial, and random. If these are wrong you just destroy your item even if everything else adds up
5. In the end, the item also has a random chance of upgrading up to 4 times. This seems to be about 10% for each upgrade. Usually if your item does not upgrade atleast once, it is too bad for you. 
6. If all of the above add upp, congratz, you just increased in power by about 1-2%. Now repeat this process for all 12 gearslots making this way harder each time. 
This system is horrible and is one of the worst systems in history of games. 

A good way to avoid all this is to offer players the means to work for their gear by hand without letting RNG dominate your progress. Giving everyone the means to actually work for a good weapon and finally getting it after 1 month is a million times more satisfying and fun than someone in your guild slotmachine create a godsword while you only got shit ones even when you play double the time. 

An RNG system makes the majority of the players angry, irritated, sad and disappointed most of the times. Sometimes, someone is lucky and he will feel 
"fu@!#ng finally I am so glad this is over" he will not feel happy. Nobody likes the RNG. 

Keep the RNG minimum and for casual stuff.  










Comments

  • No RNG in crafting

    High End Enchanting *may* have RNG.  Still undecided.   
  • A lot of what you describe sounds like an ARPG like path of exile or some other loot driven game. The game doesn't really have RNG in crafting as Jahlon said and the loot is supposed to be more sensible (e.g. wolves will drop pelt not a breastplate). Crafting will be a big part of the game and combined with gear decay, should lead to a nice healthy economy.
  • Yes I played and was a sucessful crafter in SWG,  Although nearly everything in the crafting system in SWG is the best I have seen the RNG in crafting wanted to make me cry sometimes.  One craft all amasings and a capped super weapon the next customer roles up with all their maxed out components not a single amasing and a very poor dps result for their weapon.  So glad that won't be an issue here. :)
  • The thing about RNG is that despite how annoying and stress inducing it is, when you finally get what you wanted the satisfaction is pretty big :3, well there wont be too much rng in ashes anyway :>
  • ArchivedUserArchivedUser Guest
    edited October 2018
    Like hot peppers. A little can spice things up a bit enhancing the flavor of food, but tooo much can cause you to barf, ending the date early...
  • After playing games like D&D xcom and other rng based games, I learnt that rng might love you one day and hate you other days, but when you get the perfect one in a million roll, nothing compares. The problem is making rng as fair as possible or at least reducing the risks and that's the tricky part, so I do look forward to seeing how intrepid do this  
  • The game needs some base RNG which they've said it will have, then like Jahlon said, the enchanting profession may have a bit more RNG tied in with it.

    But yeah, stay away from Warforging and Titanforging. Too much RNG is not good, but we do need some RNG for progression purposes.
  • Hi guys, ive been following this game a little bit so far but not that much. this post made me want to comment because i was wondering if the resources would have any variation to them like in SWG, or is all elm wood just elm wood ? any experimentation points on items? I was a huge huge fan of crafting in SWG. multiple accounts for an array of 40 harvesters and scouring the planets for the highest quality resources - many factories and a mini person city of crafting.
  • Says said:
    Hi guys, ive been following this game a little bit so far but not that much. this post made me want to comment because i was wondering if the resources would have any variation to them like in SWG, or is all elm wood just elm wood ? any experimentation points on items? I was a huge huge fan of crafting in SWG. multiple accounts for an array of 40 harvesters and scouring the planets for the highest quality resources - many factories and a mini person city of crafting.
    No info on that yet. I really do hope that all elm wood isn't elm wood. The wiki has some info but won't answer your question if you want to know what we DO know so far.

    This thread kinda evolved to talk about RNG stats on resources if you care to read but again, no confirmation as far as I know:
    https://forums.ashesofcreation.com/discussion/40783/crafting-resource-node-spawns
  • I think that RNG should really be a variance in quality not a success rate variance. What i mean by this is if you are crafting a club, assuming you have the skills to make said club, you can achieve anywhere from 90% of base stats to 110% of base stats. Of course the percentages can vary but it should never be an outright Succeed/Fail calculation.
    I also think all of those sub-optimal failures should be able to be reclaimed and remade. At the very least all your failures should be able to be reused in someway.
  • @Xontian
    I think they've already confirmed crafting "success" will have no RNG, you will 100% create the item. In terms of quality, not sure all the factors involved but in Krojak's stream Steven confirmed today that resources will have quality and the quality of the resources will impact the quality of the end product made from them (I posted the update on the forum post I mentioned above).
  • NeuroGuy said:
    @Xontian
    I think they've already confirmed crafting "success" will have no RNG, you will 100% create the item. In terms of quality, not sure all the factors involved but in Krojak's stream Steven confirmed today that resources will have quality and the quality of the resources will impact the quality of the end product made from them (I posted the update on the forum post I mentioned above).
    Meaning hard work will really pay off.
  • ArchivedUserArchivedUser Guest
    edited October 2018
    @NeuroGuy Can you link that Krojak interview. If it is new I definitely want to watch it.


    Edit: Thanks for the link below

  • ArchivedUserArchivedUser Guest
    edited October 2018
    @UnknownSystemError
    It's not really an interview, he just hopped in game with Krojak today (like he did with LazyPeon yesterday) and answered some questions while playing. There were more repeat questions than optimal but there were a few of us trying to answer questions we already knew the answers to in chat throughout to minimize that.

    Link: 
    Starts at ~0:55:10
  • The thing about RNG is that despite how annoying and stress inducing it is, when you finally get what you wanted the satisfaction is pretty big :3, well there wont be too much rng in ashes anyway :>
    Well this is it in a nut shell. RNG illicites a strong emotional response.
    RNG is fine if its fair and transparent, where the user can make an informed decision on if they wish to take the risk.
    Drop rates are normally broken because games do not store per player per item seed keys.
    The 1 in 100 is a lie in most games because the seed key is server based and therefore non sequential from the players perspective.
    aka BROKEN !
  • The 1 in 100 is a lie in most games because the seed key is server based and therefore non sequential from the players perspective.
    aka BROKEN !
    I'm not sure I follow this argument. Do you want 1 in 100 mean that you are guaranteed something only once in 100 events? Because 1 in 100 usually refers to 1% occurrence given an event which may practically take 2 events to occur, or 1000 (doesn't make it lower or higher than 1% in the population).
  • Thou shalt honor the stat system, your RNGesus, and generate chaos wirh no others ecxept me.
  • NeuroGuy said:
    The 1 in 100 is a lie in most games because the seed key is server based and therefore non sequential from the players perspective.
    aka BROKEN !
    I'm not sure I follow this argument. Do you want 1 in 100 mean that you are guaranteed something only once in 100 events? Because 1 in 100 usually refers to 1% occurrence given an event which may practically take 2 events to occur, or 1000 (doesn't make it lower or higher than 1% in the population).
    The way pseudoRNG is designed to work is to translate a sequence of numbers into a randomised sequence (Linear Congruential Generator). You still get 1-1000 if the pool size was 1000, but the order would change. With the same formula, sequence seed key 300 would always throw out 424 for example. You can of course change the pattern/formula after each full cycle.

    This guarantees every number from 1-1000 appears exactly once and once only when you increment the seed key from 1-1000. This guarantees a drop rate of 1:1000. You dont know when your number will appear in the randomised result order of 1-1000 keys. Just that it appears exactly once every 1000 tries.

    The problem for most games is they dont store that seed key per player. Every call to the RNG algorithm typically auto increments the seed key. That happens every time its called by any program or player. So a 100 other players may be incrementing your seed key (with server side calculations) and thus mess up 'your' seed sequence and personal drop rate. The guaranteed 1:1000 in this case has been tainted and corrupted. The 1:1000 is now invalid.

  • @Rune_Relic
    The way you claim it should be coded sounds awful to me, it would also enforce the probability over an X number of iterations (you would never get more than 1 occurrence in 100 events if the rate was 1%). This would mean if you hit your 1 event on your first try on a 1% probability, you know you have to burn through 99 more failed events until you have a chance of getting it again.

    This can't possibly be how people code RNG, you could do it in many ways where you don't need to save a seed. Could someone confirm (no offense of course)?
  • ArchivedUserArchivedUser Guest
    edited October 2018


    There are other forms of pRNG generators that dont use uniform distribution, but then you arent going to get 1 in X when someone says 1 in X.

  • I thought I would give you a more thorough explanation (without the nerd speak).
    Imagine a round clock numbered 1-100 with one rotating hand that points at one number after another.
    Pick a number on that clock which will be yours.

    1. Basic Drop Rate of 1:100.
    a.) I pick number 5. The clocks number sequence is in order. The clock ticks round and round one number after another.
    • Cycle 1 ~ my objective drop rate is 1:5 = average 20:100
    • Cycle 2 ~ my objective drop rate is 2:105 = average 1.90:100
    • Cycle 3 ~ my objective drop rate is 3:205 = average 1.46:100
    • Cycle 4 ~ my objective drop rate is 4:305 = average 1.31:100
    • Cycle 5 ~ my objective drop rate is 5:405 = average 1.23:100
    • ...
    • Cycle 100 ~ my objective drop rate is 100:9905 = average 1.00:100

    b.) I pick number 5. The clocks number sequence is in order. The clock ticks round and round one number after another....but resets every time I succeed.
    • Cycle 1 ~ my objective drop rate is 1:5 = average 20:100
    • Cycle 2 ~ my objective drop rate is 2:10 = average 20:100
    • Cycle 3 ~ my objective drop rate is 3:15 = average 20:100
    • Cycle 4 ~ my objective drop rate is 4:20 = average 20:100
    • Cycle 5 ~ my objective drop rate is 5:25 = average 20:100
    • ...
    • Cycle 100 ~ my objective drop rate is 100:500 = average 20:100

    ....because I havent used all the 100 clock numbers, I have broken the 1:100 drop rate. It doesnt matter what number I chose to be my favourite (well 100 aside), only by using all 100 numbers does the 1:100 rate remain accurate (eventually). Sure the early attempts break the 1:100 rule, but over time the error gets less and less significant.

    2. Preferential pRNG Drop Rate of 1:100.
    Essentially the same as 1b above. You dont know where in the order of the 1st 100 your preferred number exists. But the clock resets on success.
    If its 3rd 20th 60th 43rd 15th then your drop rate will be 5:141 or 3.55:100 rather than 5:415 or 1.20:100

    3. Detrimental pRNG Drop Rate of 1:100.
    This occurs when other people increment the clock passed your number (ie server side rather than client side increment) and so never succeeded when it should have done.
    If its 3rd 20th(missed=100)  60th 43rd(missed=100) 15th then your drop rate will be 3:278 or 1.08:100 rather than 5:415 or 1.20:100
    (Note I broke the preferential method and still made it detrimental. If I used 1a as the basis, then the objective drop rate would be even worse with 3:415 or 0.72:100)

    4. Balanced pRNG Drop Rate of 1:100.
    Provided you dont truncate the total counter or miss your numbers from those cycles, balance pRNG gives you the same overall success rate of 1:100, but randomises the early success rate.
    If its 3rd 20th 60th 43rd 15th then...
    Cycle 1 ~ my objective drop rate is 1:3 = average 33.33:100
    Cycle 2 ~ my objective drop rate is 2:120 = average 1.67:100
    Cycle 3 ~ my objective drop rate is 3:260 = average 1.15:100
    Cycle 4 ~ my objective drop rate is 4:343 = average 1.17:100
    Cycle 5 ~ my objective drop rate is 5:415 = average 1.20:100

    TL;DR
    A Drop rate of 1:100 must mean an objective drop rate of 1:100 no matter how long and hard you grind for.
    Whether you get drops on 100 & 1 with no gap, or 1 and 100 with 199 gap is neither here nor there, as long as it averages 1:100 over time.

  • @Rune_Relic
    Man, thanks for putting in the effort to explain that but it's not that I don't understand how the pRNG you suggest works. I'm actually a programmer and work with pRNG on a daily (a version of twister, which btw I thought twister was pretty much standard now). I just don't know much about MMOs and sending and encrypting data to servers and whatever else is involved in an online game.

    I understood what you were saying I just am skeptical that the system you describe is what people intend and try to implement. It requires so much data to be saved for each RNG calculation... you need to save the probability, the vector, your place in the vector. It seems inefficient and in most games a 20% chance event can occur 3 times in a row and it's a feature not a bug. When you take issue with how pRNG is implemented elsewhere you are considering things for an individual case for a finite amount of time. It's like only flipping a coin 5 times, of course you won't get 50:50 but that doesn't change the fact that heads and tails are 50:50. Similarly if you flip a coin 5000000 times, you may get a streak of 100 heads in a row, that still doesn't change the fact that tails and heads are 50:50. Your system would constrain a heads-tails-heads-tails cycle which is strange within the context of RNG in game (I've never seen that, maybe you could give an example?).

    In any case, appreciate the time you put into that budd. I just would need corroboration of your claim I hope you understand that. Cheers! :)
  • ArchivedUserArchivedUser Guest
    edited October 2018
    Any RNG mechanics especially with progression or crafting is a deal breaker.
    After BDO me and my friends will NEVER play a MMO again that has RNG mechanics for progression or anything other then regular loot drops.

    And if i am going to pay monthly for the service any Pay2Win or RNG to progress is just a cash grab and unacceptable IMHO..

    But i will gladly then spend money on mount skins and costumes and so 4th..
    As long as am not forced to because there is no visual progression in-game.
  • @isfjell
    You've come to the right game then my friend. Virtually no RNG in crafting (small exception of "over-enchanting" where you take a calculated risk to put on extra-enchantments that may destroy the item). No RNG in progression (not sure what that even looks like tbh). And there is non-skin visual progression.
  • NeuroGuy said:
    @isfjell
    You've come to the right game then my friend. Virtually no RNG in crafting (small exception of "over-enchanting" where you take a calculated risk to put on extra-enchantments that may destroy the item). No RNG in progression (not sure what that even looks like tbh). And there is non-skin visual progression.
    Tera . . . RNG when enchanting gear
  • Im banned from looting in my guild in raids due to my bad RNG luck XD
Sign In or Register to comment.