The site's dice are accurately pseudorandom.

I would really like to see someone complaining of having good dice. Why nobody is giving examples of killing 20 and losing 1-2? [quote]i play 3-4 games a week. with regular dice and i can tell you that these dice do not emulate real life. normally the attacker has a slight advantage. i have never seen that over the long haul that that is what happens.. im not a programmer but as i said i play with real dice every week...and these do NOT act like real dice.[/quote] @@kwikool do you know that here you can do 100 attacks in just few seconds but in real live you will spend hours doing them? I have played risk in real life as well and you cannot compare the real life risk with this one, at least the risks I was playing were with a small amount of units like playing on classic map where from cards you are taking from 4 to 12 units and nothing more. And it's really rare and hard to stack more than 15-20 units in one territory and to attack a large amount of units so you can experience the real randomness. While on the other hand here you are playing on increasing cards (probably) and after half an hour you have received and wasted more than a 100 troops each of you in the game and it's normal to experience big difference in kills and losses. Also it's NOT SAME to attack with 20 troops against 10 units in 1 territory and attacking with 20 against 10 in for example 6-7 territories. You have a pretty higher chance of losing in the second example than in the first so calculate your attacks better to see what's the odds of losing and winning the battle.

[quote=God_of_War]... you can't do anything ... No matter how much you "switch it up" wait, attack other areas, etc... [/quote] That's because it doesn't make a difference. Randomness is randomness. All that 'switch it up' and 'click with my left hand' and 'wait for full moon on sundays' is just superstition. [quote=God_of_War]... It simply feels ... I believe you, but you also have to believe us too. :) [/quote] The thing is, I know that it feels this way sometimes, and I've felt myself plenty of times this way as well. The same things that make the game so interesting (randomness and human interaction), are also what make it really frustrating at times. One thing that makes the frustration less is using the balanced dice - you're actively making it [i]less[/i] like the real dice that way, but IMO that's better. The other thing that helps is the realisation that it's a game, and even if you play perfectly, you can't win every time. [size=3px]Just like real life actually.[/size]

[quote=God_of_War]Run the test of 45 vs 20 and let me know how many times you get kill only 15 and die 45. That' what put me over the edge. [/quote] Well, first thing first, you cannot lose more than 44 when attacking with 45. Second, I assume that when you say "killing only 15" you mean "15 or less" ( not "exactly 15" ) so I count also when one kills only 14 or 13 or even less, down to nothing. So summing together all those cases you have a 0.01231% chance of losing 44, killing only 15 or less. It's like saying that it happens (in average) once every 8,123 tries... Rare of course, but much less rare than winning the national lottery, and still somone wins it every year. Once every 8,000 does not mean that it never happens: it just means you got really unlucky. @@dimceto ,, I agree with you. Just let me specify one thing: Americans play increasing cards on the real boards. Yes, an attack like 120 v 100 can take very long, but usually people suicide before getting there! :D The flate rate cards (4 to 12) is just a European variant of the original Unitedstatser Risk.

@@The_Bishop thanks for the info I didn't know that, but still when playing with friends as you said most of the times someone suicides so still pretty unreal to compare real risk with this one here

I think that in Risk as in life there are 2 different approaches. Some people like to talk about their problems and misfortunes or whatever has gone wrong in their daily life. Others don't like talking about these things and just want to forget and move on. The former are more frustrated but also have a greater ability to react to situations; the latter, on the other hand, have a more relaxed life but sometimes accept things a little passively. So, those who complain about the dice here ... I don't think they're really saying the dice are wrong -- also because they know that tests can be run (and have been) proving the fact that the dice work perfectly as intended -- they just want to have the pleasure of coming here and telling everyone that they have been unlucky, and not just once, but several times! :o Don't worry guys: these things are normal and happen to everyone. Just not everyone thinks it makes sense to come in the forums and type "I killed 1 and lost 15" because ... First, who cares? And second, what gives? But okay, if you can't resist the need to report your bad rolls, then go ahead, do it.

Curious Bishop...when you say you have run tests, what tests have you run them against? Did you do simulations using D12 dice? What are you comparing them to? Because when I ran my tests using a few varieties of algorithms, the original that D12 used to use was more in line with what was expected.

Well, I saw your statistics and those are interesting and very accurate. I think you should repeat them (please). But still they report only a small difference from the expected value, 100 on 40,000 is just 0,25%, I don't think it's enough to get the percection of it. So okay, our virtual dice perhaps aren't perfect as I said, but almost perfect. My tests (I told in brief) were based on sets of 10 rolls, each roll being 3 dice vs 2 dice. Because as you can read, people have a sensation like the dice get stuck and keep giving a losing (or winning) outcome several times in a raw. If that really happens we should observe an anomalous distribution of the outcomes of 10 rolls. Instead I found all normal, both the average and the deviation. I did it manually using two accounts, creating test games in which I piled up hundreds of troops in several territories. Quite a long process, so I couldn't reach a very high accuracy, but at least I tested the virtual dice in the real place they are used. It was years ago, so probably the random function used was [i]mt_rand[/i]. At some point I got bored and stopped the test, maybe I will run it again, one day, but really it's time consuming... I like a lot the solution found by GamesByEmail, http://gamesbyemail.com/DiceGenerator. Just a [b]real random[/b] system. But I think we need a lot more than one die roll per second, which their machine can provide. And 30 or 40 machines would require ample space, supervisors / technicians, and a lot of electricity costs.

One frustration I've noticed is that I have regular instances where an attack fails even with 99+% chance of success. The frustration is that I never have success with attacks that have <1% chance of success. But of course that's because I'd have to be an idiot to try that kind of attack. So the perception can be skewed that way. That said, I've had plenty of instances, especially when attacking a large stack with a large stack where the dice give me <1% chance results in my favor. Sometimes an attack might start off killing 20 and only losing 2 (or that might happen in the middle of the attack, or at the end). But then as the attack continues it tends to even out to what's expected. I did have one recent instance when I was just setting up for my next kill right before turning in a set, so I attacked 34 with 12 just to see if I could weaken that territory a little and I ended up killing all 34 and only lost 6.

mt_rand is no longer used. In my tests it was more closely aligned with the %'s provided on expected outcome.

[quote=B4rny]U would think, attacker has the advantage, but I'm thinking that's not the case[/quote] Correct, the defence has the advantage, no matter how many dice the attacker uses, all the defence has to do is, match the highest 1 or 2 dice. This may be controversial but, it seems to be something that is overlooked and should be formulated and factored in. Attacker & Defender have an equal chance of rolling any of the 6 numbers on the same amount of dice, however, the value of the result is different for both. What is to stop the defence dice from matching or beating every single roll of the attack?

Ehhh, no? That is not correct? In a 3 dice vs 2 dice the attacker has the statistical advantage. The advantage is only small, due to the equal dice advantage for the defender that you mentioned, but the extra dice advantage helps more than that the equal dice advantage does. Of course, a statistical advantage doesn't mean you will always win, or even that you'll win in most cases (one can be unlucky). Just that it's more likely to win giving equal amount of troops (in large amounts).