paying customers have battle advantages

[76561199801620088]

its the only way to explain the 5 v 2 or 7 vs 2 loses that happen often

[FlappyJak]

7v2 is a 100% win in Balanced blitz (so is 6v2). So I assume you're talking about True Random. Which can can be very, *unsurprisingly*, random. TR favours more the defender in smaller battles than Balanced Blitz does. It's a common tactic in TR to leave 2 troops in every territory, as this greatly increases your chances of killing attacking troops versus 1.

It's got nothing to do with paying or not paying. It's just bad luck.

[Kenpoleon Bonaparte]

Originally posted by FlappyJak:

So I assume you're talking about True Random. Which can can be very, *unsurprisingly*, random.

The problem is that the label "True Random" is neither true, nor random. It would be more accurate to call it "True Even", because what it does it even out your total dice rolls so that each number on the dice is rolled 16% of the time.

https://steamcommunity.com/sharedfiles/filedetails/?id=3474991199

[FlappyJak]

Over a large enough data set your rolls would even out, absolutely. This is true for any amount of possible outcomes with an even chance and enough attempts. But the 16% you mention is only counting how often you roll numbers, not how often those numbers actually help you win. You could attack with 6, 6, 6, if the defender rolls, 6, 6 you still lose, even though you got more 6s!

The problem is perception. People have a few bad, maybe critical to the game, rolls. And they remember those and deem them unfair. But it's still random, and essential to the game of Risk. I've absolutely suffered to bad dice before, but I don't jump to the conclusion that it must just be rigged against me.

To add. If your rolls were NOT even. Then that'd show it WASN'T random. But they are, so it is.

[Kenpoleon Bonaparte]

Originally posted by FlappyJak:

Over a large enough data set your rolls would even out, absolutely.

Wrong. Studies on winning lottery numbers show that certain numbers come up more than others. For example, the reason why 25 has been picked 8 times in the last year, while 14 has only been picked 2 times... is because it's random. All numbers coming up the exact same number of times is a pattern, and by definition a pattern is not random.

Yes, we've all suffered greatly at times from bad dice rolls. It IS part of the game.

[FlappyJak]

Actually it's mathematically a fact. What you're noting there is randomness bias. Like I say, it's about the size of possible outcomes, and the number of attempts. With dice rolls in this game, we're talking merely about 6 possible outcomes over a lot of attempts, which averages out to 16% per outcome. The lottery is a large amount of outcomes (1 - 59) over relatively far fewer attempts. Given a lot more attempts the numbers would also average out to be even. That they haven't, merely shows there's not enough attempts.

[Kenpoleon Bonaparte]

Originally posted by FlappyJak:

The lottery is a large amount of outcomes (1 - 59) over relatively far fewer attempts. Given a lot more attempts the numbers would also average out to be even. That they haven't, merely shows there's not enough attempts.

Even something as simple as the 'Daily Pick 3', which I think now is played 4 times per day, doesn't have perfectly even numbers. And lotteries have been legal for several decades, so there is plenty of data to study. Most lotteries have their histories online now. The 'Daily Pick 3' has ten numbers, so by your logic the records should show that each number is equally picked 10% of the time. Look up any long-term time frame you want, and that is not the results you will see.

There is a difference between "possibility" and "probability", which is something most people don't understand. Each number has the possibility, or equal chance, of getting picked versus any other number. However, each number doesn't have the same probability, because certain numbers do in fact get picked more. That's the randomness of it all.

[FlappyJak]

As I said, mathematically; any number of outcomes of even chance, after enough tries, will all even out. If the results are not even, it's either not truly random, or not enough tries have been attempted.

The number of tries is very important. And I feel when you quote lottery stats this is the point not being factored in. What you're pointing out is anomalies. Anomalies tend to disappear with more tries. To say "25 has been pulled 8 times this year, 14 has been pulled only 2 times", is not the same as "25 will always be pulled 4 times as often as 14.".

Random number generators are tested by this exact nature. A good RNG will eventually even out (like Risk 16% across the board, this is a sign the RNG is GOOD). A bad RNG; will often avoid the lowest/highest numbers, or have a tendency towards a certain value (imagine a shallow bell curve, etc).

Also worth noting. The randomness of sets of values is a lot different than individual numbers. If you were to measure the results of 2d6, you'd have a bell curve centred over 7. But that even this is a results of the individual rolls being even.

EDIT: Just wanted to come back and share this with you: https://anydice.com/
There was a better tester somewhere that shows the rolls over time, letting you see the anomalies but I can't recall where that was.

[Jim]

I've played paid and free a few thousand times,
I've havn't seen any difference in the dice rolls or the bots behavior in FFA.
Ps...I've never lost a 7-2...and the 5 on 2 doesn't always work for me either.

[DreamSword Souty]

Originally posted by Kenpoleon Bonaparte:

Originally posted by FlappyJak:

So I assume you're talking about True Random. Which can can be very, *unsurprisingly*, random.

The problem is that the label "True Random" is neither true, nor random. It would be more accurate to call it "True Even", because what it does it even out your total dice rolls so that each number on the dice is rolled 16% of the time.

https://steamcommunity.com/sharedfiles/filedetails/?id=3474991199

Thanks for sharing this information, I didn't know and I learned something!