Talk Fantasy Football

dode74 wrote:

Everything with the RNG is predetermined, which is why they use massive numbers for a 1-6 result. Dice are not.. without fixing the roll in some way (shave/weight dice, short-drop roll) you can't predict dice better than the odds themselves.. unless a die literally breaks off a corner during a roll, one roll does not affect the result of the next.

You can't predict the RNG better than the odds themselves either. What's the difference?

If your only interest is trying to predict a number between one and six, none.

The original statement was not "the only difference between guessing the next 1-6 number generated is the RNG and the dice"..

If you are hungry, and get fed, does it matter what the food is or how it is prepared?

Let me put it another way.. If I play for fun, I'd rather use dice and tabletop if given the choice, it is (to me) more fun .. if I play for money, I'd rather have an actual randomly generated result at the time than trust in the sequence the computer has already set up. Again, thats a preference.

Would you play the lottery if you knew the numbers were already drawn 2 days ago?

Smeborg wrote:Way to go... off topic.

\o/ Happy to help with that. Especially with this dreadful Claw discussion that won't end.

Daht wrote:If your only interest is trying to predict a number between one and six, none.

The original statement was not "the only difference between guessing the next 1-6 number generated is the RNG and the dice"..

If you are hungry, and get fed, does it matter what the food is or how it is prepared?

Let me put it another way.. If I play for fun, I'd rather use dice and tabletop if given the choice, it is (to me) more fun .. if I play for money, I'd rather have an actual randomly generated result at the time than trust in the sequence the computer has already set up. Again, thats a preference.

So, to be totally clear, there is no real difference, just a perceived one?

Would you play the lottery if you knew the numbers were already drawn 2 days ago?

Sure, why not?

"So, to be totally clear, there is no real difference, just a perceived one?"

Same way there is no real difference in playing with paper markers and miniatures, only percieved ones.

Same as there is no real difference walking or driving to work, just a perceived one.

No real difference in just spitting out a streak of numbers between 1 and 6 200 times, and playing a game of blood bowl.

The end result is the same, therefore they are the same. There is no real difference between using a RNG and getting a 6, rolling a dice and getting a 6, or placing a dice down with the 6 facing up.

No, there are real differences between those things.

The perceptual difference of how a random number is generated when all that matters is generating the random number (the process is irrelevant since the process is not a part of the game) is very different to the real difference loonie was talking about.

Ah, I see.. you didn't realize from the start what loonie's point was...

An RNG is not a programmed dice roll. It's a math trick to simulate it. Thats why you use 4 billion results to get a number between 1 and 6. If you could program a dice roll of 1-6 why would you waste the cpu effort to generate 4 billion results?

Frankly if cyanide were to put any programming effort into their game it should be in the AI not the rng anyway...


If you remove how and why from an equation, you are correct. (Which also applies to each of my examples in the previous post.. getting to work is getting to work, eating is eating, a string of numbers is a string of numbers)

You are arguing generation of numbers for no purpose and in no context, under the pretense that all randomness is merely a complex certainty impossible to predict on the fly.... I'm arguing about blood bowl, the hows and whys of things, and of the doing of things.

We're both right, in our own context.

Thanks, but I know how the RNG works and why you can't simulate a dice
It takes no effort to produce 4bn results, btw, since you actually do produce them on the fly. All you do is apply an algorithm to a seed to get the next result and seed, then rinse and repeat. The complexity of the set of 4bn results is all stored in the algorithm, just as the algorithm y=x^2 stores the complexity of the set of all possible positive numbers - neither one takes CPU power until it is needed to produce the next result.

I'm not talking about removing the how and the why, I'm talking about "fit for purpose", which is what we need to achieve the aim. Walking to work may not be fit for purpose due to other effects (time taken, for example), paper markers may not be fit for purpose (GW rules for tournaments, and also the aesthetic desire to use well-painted minis).
In terms of the RNG, fit for purpose is defined (by me, at least) as "can it produce an unpredictable number from 1 to 6 with a good approximation of a 1/6 chance of each for the duration of a game," and the answer for this RNG is absolutely yes - far better than GW dice, in fact! If you have another "fit for purpose" definition for an RNG then I'd love to hear it.

Ooh, stochastic v epistemological probability. Love it. I flip a coin, does it make any difference to you whether you call when it is in the air (stochastic, random) or once it is covered in my hand (epistemological, you just don't know)? Try to teach the kids this every now and again - cue lots of blank looks

PS now, about claw...

It's all epistomological - it's just a lack of data and/or processing power within the right timeframe. Right down untill you start hitting Heisenberg (or perhaps wave functions).

Which reminds me, apparently there's a hotel in Germany where there's a plaque which says "Heisenberg may have slept here on 21 Jun 1925"....

True, unless humans have free will

Thumbing dice and palming coins is generally considered cheating though

The argument was never if it's good enough to fit the purpose, it was if there was no difference.

That entirely depends on if how and why matter. Both of your counters to my example assume the how and why matter, where your argument on the RNG depends on the opposite.

If someone could actually program a die roll, they would (it would be the closest to cheat-proof system ever made).. so they do the next best thing.

How much the how and why matters depends on the player. I like pushing my luck and playing hyper-aggressive defence (and my league record shows it ) ... to me there is a definite difference in the feel and flow of the game between 'is the next predetermined number high enough' and 'can I make this roll' ... to me a BIG part of the game is the social/tactile aspect, and the hobby.. I like online also but it's strictly a substitute. Honsestly to me even IF you could program an honest 1-6 die roll, it would only cover half the difference between RNG and an actual dice.

To me, all of this is also in the context of a game played strictly for fun. Taking the how and why out of that makes the whole argument completely meaningless.

Actually the question was what was meant by "The MAJOR difference of TT v/s computer is the dice". If all we're talking about is that the difference is that it is simulated then that applies to the entire experience. According to your own criteria you can never properly simulate the experience, and I should imagine that for you the main difference is that you don't actually face an opponent.

So again I ask for clarification on loonie's statement (the original question) - why is the fact that a PRNG used the "MAJOR difference" between the two versions? The PRNG is as good as dice in terms of fitness for purpose, just as the pixels are as good as minis in terms of fitness for purpose (for playing, and in many cases for "painting"). Any difference in the fact that a particular element is simulated applies across the board.

Let's say we generate 1000 rolls with computer and get the average. It should be pretty close to 3.5, right? But if we roll 1000 actual dice rolls, it may not be the "real" average, because it is so called true random and it does what it wants. It could be 3.45 or something.

Did I get it right?

You will only average exactly 3.5 with the RNG if you cover the entire 4bn number series, or are very lucky with your start point. In fact, you'd have to be as lucky with the start point as you would need to be to roll that same average with real dice!