RNG as a concept: Discuss

#1 - Oct. 15, 2014, 1:27 p.m.

This is a spinoff of the economy thread to talk about RNG tactics in games in a general form.
Here’s the premise. RNG is evenly distributed on aggregate. On an individual level this means that while almost everyone falls into a reasonable range in the middle, there are outliers on each side of the distribution that are either highly rewarded or not rewarded at all. These individuals become sample cases and spotlights for experiences that maybe shouldn’t exist.

We do need to be very careful about ideas that flatten the experience entirely as that quickly becomes not fun at all.

There are two concepts that have been discussed in the other thread that I’ll briefly summarize.

1. Use a specifically non-random NG. The NRNG functions similarly to a RNG, but has characteristics that either squish the distribution so that outliers exist much less or specifically manipulate a player’s experience for loot in a more complicated way that makes it feel rewarding.

2. Implement measures that counteract low-end outlier behavior inside of game design. This would be a system that is something like: If player hasn’t received a rare drop in X time send them Y tickets for random drops.

2.5: “Add secondary reward mechanisms (ie. token based system) alongside the primary RNG system; allow progress to be made even when you don’t get the result you want.”

Obviously these are hyper-simplified descriptions, but I don’t want this to get too long.

edit: added 2.5

#69 - Oct. 15, 2014, 4:25 p.m.

munkiman.3068:

phys.7689:

Brother Grimm.5176:

…
I’m pretty sure that nobody think there is special code about that.

But we’re a lot to rather think that RNG is bugged in some way, and that’s why, in addition to a “gauss distribution” as described in the previous post, some players are always lucky and some are always unlucky. As the RNG mechanism is all in all quite complex (modified by magic find ; depends on damage done by the player / by his group compared to damage done by other players ; special items with a different table loot ; …), there could be several bugs introduced in the system. Remember that the % of magic find displayed in the hero panel is wrong since months (but the real number is supposed to be correctly applyed), so it wouldn’t be surprising that there are some bugs in the loot mechanism itself.

There are CERTAINLY very vocal players that are convinced of the “secret LUCK stat”.

its not really relevant weather there is a secret luck stat or not. If a random works properly, there will be people who will, will seem lucky, or those who seem unlucky, when compared to the data. This is the normal expectation of the bell curve.

So whether it is a bug or not doesnt matter. The point is these people are bound to exist.

My take on RNG, isn’t that it exist or it’s broken, it’s just that it’s practically everywhere. Neither of those solutions would work anyway, since the game would have to predict what you’re after in the first place, then reward you for failure of what exactly? Do we put a merchant in the game that carries everything a player might want in exchange for a loser token?

This was the first problem I thought of as well. It’s simple to say failure and success, but that’s not an easily definable concept in real life.

#204 - Oct. 16, 2014, 1:24 p.m.

phys.7689:

Wanze.8410:

Nevets Crimsonwing.5271:

I’ve not read this whole thread so forgive me if this is answered elsewhere. I want to ask a few questions.

1. Is the loot system currently set up so people who play less are more likely to get an awesome RNG drop? There have been numerous posts like, “my friend logged on after 5 months and got X, Y, Z drop while I’ve been playing every day and got nothing!” I want to know if the way it works currently is that those who play lots get decreasing loot in proportion to playtime. Why would ANET do this? To keep people chasing the carrot.

John answered that question in a different thread, there arent any account modifiers.

https://forum-en.guildwars2.com/forum/game/bltc/I-have-a-question-about-the-economy/page/15#post4487484

what he is speaking of wouldnt be an account modifier, and it probably does exist, mildy. There is Diminishing returns on drops and it decays pretty slowly. You will in fact see more drops in your first few hours after not having played for awhile.

They will never tell us exactly what is happening, but your friend who plays rarely will basically have no DR for almost all of their playtime.

Still catching up, but I wanted to stop to be clear. No to this. This doesn’t happen, it’s a cognitive bias. There’s virtually no DR active on any given day.

#210 - Oct. 16, 2014, 2:33 p.m.

Cakemeister.5792:

I just want magic find to make a difference.

Magic find makes a huge difference.

#402 - Oct. 20, 2014, 2:22 p.m.

Checking back in. This isn’t meant to be a discussion of just those three ideas that I posted, those were just a primer for options. The idea is to discuss RNG as a concept inside of games and reward systems. I think more discussion will lead to better understanding of the problems current systems may have, and the pros and cons of other possible systems.

That being said, this is not a discussion of how you should get a precursor. I’d prefer the discussion to be much more general, solving specific problems is easy once you have an excellent framework.

#421 - Oct. 20, 2014, 6:46 p.m.

Marcus Greythorne.6843:

John Smith.4610:

The idea is to discuss RNG as a concept inside of games and reward systems. I think more discussion will lead to better understanding of the problems current systems may have, and the pros and cons of other possible systems.

John, could you give us a bit direction here? I’m not quite sure what specifics we could/should discuss… it seems pretty one-sided that many think RNG is bad and tokens could work.

At the heart, there’s 3 basic options:
1. Mostly RNG
2. No RNG
3. Hyrbid RNG (where most games are)

starter concepts for cons of each system:

1. Many players can feel like they have “bad luck”, in fact players with standard to pretty good luck are still likely to feel bad because of how humans interpret data.

2. A completely predictable experience lacks moments when something fantastic happens.

3. You run the risk of getting the cons of both systems without very precise design

An important question may be to ask if a model is still a good concept and a great implementation just needs to be focused on or if the model doesn’t work fundamentally.

#434 - Oct. 20, 2014, 11:09 p.m.

Essence Snow.3194:

John Smith.4610:

At the heart, there’s 3 basic options:
1. Mostly RNG
2. No RNG
3. Hyrbid RNG (where most games are)

starter concepts for cons of each system:

1. Many players can feel like they have “bad luck”, in fact players with standard to pretty good luck are still likely to feel bad because of how humans interpret data.

2. A completely predictable experience lacks moments when something fantastic happens.

3. You run the risk of getting the cons of both systems without very precise design

An important question may be to ask if a model is still a good concept and a great implementation just needs to be focused on or if the model doesn’t work fundamentally.

I really have to comment on leading in both this post and the OP. By all means post, but please, please, please leave the leading biases out if the intent is to have a productive honest discussion.

edit……I’ll give you an example of what I am referring

Which would you rather have
1) A savory treat which leaves you wanting more
2)A sweet treat which puts on the pounds
3)A sour treat which tests your bravado

The biases are in the 2nd and 3rd. The former being negative and the latter being positive. Given this leading most are apt to choose option 3 and less likely to admit to 2.

Both the OP and this post are reiterations of what’s been said already, I’m not adding new or personal content to the discussion yet.