RNG-itis: a cure without (notably) changing probability

"

aimlessgun wrote:

IMO: With fusings, you'd use a multiple deck approach, only apply it to 6l, and the objective would be to make the 6l more likely not for people simply on the wrong side of the median, but the unluckiest 1% or so of people.

It's hard for people to really grasp what that kind of unlucky even looks like for situations like this.

The formula for it is x=ln(1-m)/ln(1-p), where x is the number of orbs used, m is the percentage of unluckiness (so 0.99 in this case), and p is the chance of success each trial. If Fusings have a 0.2% (1 in 500) chance of success, that would mean spending 2300 Orbs.

"

ScrotieMcB wrote:

"

aimlessgun wrote:

IMO: With fusings, you'd use a multiple deck approach, only apply it to 6l, and the objective would be to make the 6l more likely not for people simply on the wrong side of the median, but the unluckiest 1% or so of people.

It's hard for people to really grasp what that kind of unlucky even looks like for situations like this.

The formula for it is x=ln(1-m)/ln(1-p), where x is the number of orbs used, m is the percentage of unluckiness (so 0.99 in this case), and p is the chance of success each trial. If Fusings have a 0.2% (1 in 500) chance of success, that would mean spending 2300 Orbs.

Honestly if the simplest way to fix this problem is to just implement entropy

This means for every 500 fusings, you are guaranteed a 6L. Not sure when, but you will get that 6L for every 500 fusings

Since its a guarantee they can increase the price a bit (i.e. lower the chance to get 6L) a bit, but its a much better solution to the problem

"

deteego wrote:

Honestly if the simplest way to fix this problem is to just implement entropy

If it starts in a random place, every time I get 6L with a character I'm rerolling and never crafting with him again

If it starts in a predetermined place, RNG is destroyed completely.

Amen to it being killed if it starts in a pre-determined place. Within a week you'll have guides telling you that if you start playing at 16:49:31, your 12th fusing will give you a 6L, and crap like that.

"

TL;DR: OP's suggestion would likely lead to one or both of the following:
* A general buff to all drops, which could be handled more directly with the current system
* A system with lots of calculation overhead and weird tracking rules that would essentially steal from the lucky to benefit the unlucky.

Hi, Scrotie,

In no way was I proposing reducing the luck of the lucky. Furthermore, to meaningfully reduce the probability of long poor runs for the unlucky, you would need to increase drops quite a lot. While indeed you did capture that the proposed system could have a marginal increase on drops overall, I noted in an aside that you could keep a history (over lifetime, which I think I did not say), in which case that's likely to amortize away pretty quickly. The point of the latter would be almost entirely about first impressions. Tell me if you see what I am getting at, there.

The TL;DR I would use to summarize the article is this: "Raising drops globally creates a lot of fat dogs. Have we considered the possibility that it might be better to throw the occasional starving dog a bone, instead of feeding all dogs more beef?"

Do you see the way of it?

I only bothered with the OP because I've seen this whinery before, most notably over in the D3 fora, where basically the collective community complained and complained and complained, until they successfully browbeat blizzard into giving them more stuff. But I always noticed that a lot of the arguments centered around people that declared things like '150 hours of play and not a single legendary.' I always secretly suspected that they indeed did have a legendary, but missed it on the map in all the chaos (which as you probably know, they eventually made GUI helpers for that).

As for the simple alternative you propose, "let's just raise all drops," I actually wouldn't say I oppose that if the view of the open beta community was such that drops aren't frequent enough over all. I view that as a wholly separate issue, however.

With gear specifically, I suspect that we don't need more drops, FYI; I suspect that the actual 'problem' is tied pretty closely to the HP problem currently; to wit: my own gear would be pretty stellar if the HP values were higher; I'm not saying that such should be buffed, but rather suspect that GGG is working on the A3 difficulty issue as we speak.

With currency, I have no opinion. From an economy perspective, doubling drops will halve the value of all currency, and have some consequence on items that I can't foresee. Only the latter would trouble me. The prior would be a total wash: I am indifferent.

There's also an open question of when it is appropriate to use randomness in a game like this. It might be, for example, to set it up so that 500 (or whatever#, I don't know the intended odds) fusings just makes a 6L.

This metaphor for rich and poor is maladroit. The proposed statistical change is intended to address aberrations in an environment where being "poor" is actually relatively unlikely. Tweaking something that by definition does not happen very often at all can have little statistical impact to the whole system; I in fact doubt it would be an observable at all.

"

Undon3 wrote:

...Also, breaking your gear should not happen. I'd propose a rare-ish "Orb of Remembrance" (or something), that when applied on an item, it would remember its parameters, and the item would gain an additional property, something like "Ctrl+Shift+Click to discard all changes made to the original item". Then you'd use your fusings etc., and if things work out badly, you would just get back the original item. You already lost the orbs, no need to be punished for what is not your fault (bad RNG) with an item too.
However, this orb would be but a start.

I like this idea very much, I can live with disappointment but totally breaking gear on top of losing all the crafting mats = ragequit.

I like the idea of the OP very much. Especially for jorbs and fusings. As it is now, the pure rng nature of this system causes way too much frustration to the player base.

For mods on gear, the pure RNG is good. But for something as essential to a character as socketing your gems, it's just not good.

"

deteego wrote:

"

anubite wrote:

Er, huh? Why are you telling them how to change RNG when you don't supply a reason to do it in the first place.

Evasion was changed because "randomly getting one shot" is not fun and a negative aspect so strong people did not want to use the defense stat in HC.
But randomness is otherwise, usually fun. There are times where you can get 4 uniques in a day, and times when you get nothing in 2 hours. It is these spurts and dry-spells that spice up a random game like POE.

If you make all RNG in the game deterministic, you just homogenize the experience.

Not really, if something is a really low chance than its still going to be a really low chance

For example, I heavily believe that both sockets (fusings) and colors (chromatics) should use the same entropy system that evasion does, mainly because those aspects of items actually influence your skill build, and that shouldn't be subject to that much RNG

This x10000000.

It's one thing to complain about gear drop RNG, or currency drop RNG, or item stats RNG, those things should be expected in an ARPG; but the RNG on socket color and links is quite simply stupid, due to how negatively it impacts just playing and having fun with your character by making good gem combos hard as hell to achieve without ASSLOADS of luck or fusings. Keep RNG in every other aspect of this game the same way it is for all I care; but something about chroms/fusings (especially fusings) needs to change.

It is actually pretty simple for developers/programmers to implement a mechanism (that exists only behind the curtain , of course) that would guarantee that player receives at least 1 unique during set amount of time (say 1 unique per 20 hours of /played). The reasons they are not doing this are purely design decisions not technical ones.

"

that would guarantee that player receives at least 1 unique during set amount of time

... solution ...

Leave char logged in while at work, with one of those little mouse twirler programs making the char run in circles. :-P

Pretty soon you'll say, "oh, I meant really playing," and soon thereafter you'll conclude that "random number draw count" is the best way to keep track. And voila. Back to OP. :-P