Dice analyzer

[oggiss]

Can someone tell where I can see the stats? :p

[dividedbyzero]

Can someone tell where I can see the stats? :p

There should be a menu option under the Personal Menu on the left that says "Dice Analyzer".

Click that and it pops up a stats window.


(This of course assumes you have greasemonkey and have installed the analyzer. )

[oggiss]

Ok thanks.


I have a hard time seeing if it's good or not, but with the thought of not having a win in like 3-4 games. f*ck THIS COCKGAME!!!!!!!!!!!!!!!!!!!!!!!!!!

[oggiss]

Shit, I am in a bad mood -_-

[kalishnikov]

Ok thanks.


I have a hard time seeing if it's good or not, but with the thought of not having a win in like 3-4 games. f*ck THIS COCKGAME!!!!!!!!!!!!!!!!!!!!!!!!!!

lol everyone who plays CC says that, or something similar, at least once a week.

Oh and I recommend ignoring the analyzer unless you wanna get more pissed....

[dividedbyzero]

Ok thanks.


I have a hard time seeing if it's good or not, but with the thought of not having a win in like 3-4 games.

Dude. Relax.

Pretty much everyone that's been here for any length of time has gone on a big points and game losing streak. It happens.

I did something like an 800 point nosedive at one point. I think Fitz did around 1000 points.

It just happens. Frankly, 3-4 games without a win doesn't seem *that* bad...but no one enjoys losing, I know.

[oggiss]

Ok thanks.


I have a hard time seeing if it's good or not, but with the thought of not having a win in like 3-4 games.

Dude. Relax.

Pretty much everyone that's been here for any length of time has gone on a big points and game losing streak. It happens.

I did something like an 800 point nosedive at one point. I think Fitz did around 1000 points.

It just happens. Frankly, 3-4 games without a win doesn't seem *that* bad...but no one enjoys losing, I know.

ehhh? I was playing 1 vs 1, which I have more than 65% win percentage on ^^

And besides, have you guys had this dice since you became Premium member? ^^ f*ck this

[dividedbyzero]

ehhh? I was playing 1 vs 1, which I have more than 65% win percentage on ^^

And besides, have you guys had this dice since you became Premium member? ^^ f*ck this

I've been a premium member for awhile...less than some others on the site, but perhaps a year and a half, I think. And the dice have pretty much been the same for me from the start. There are days I can't lose and days I can't win. And the can't win dice always seem to be at the worst possible time.

When the dice start going bad, I have to remember to just stop rolling and come back later. That usually helps for some unknown reason.

I find 1v1 to have more of a luck aspect than other numbers of players. Not to say that there is no skill, but the drop and dice dictate the game in most cases, I think.

At any rate, it's a cold comfort, but I've rolled just shy of 194,000 attacker dice since I last reset the analyzer and my dice are almost statistically perfect. (16.55, 16.59, 16.74, 16.67, 16.68, 16.77) for 1s to 6s respectively. My defender dice are pretty much the same for about 101,500 rolls.

I guess what keeps me from getting pissy about the streaks is that I know they happen to everyone.

[Fruitcake]

Quite frankly the dice stats here are absoloute rubbish.

I see it mentions a win rate of 67% when attacking 3 dice on 1. Then how is it my records show around 29%....

How is it my win rate when 3 v 2 is around 13%

BTW, the above is from a statistical sample of well over 100 games.

Absoloute stuff and nonsense, all of it.

[AndyDufresne]

I agree, the dice in this universe just don't work. I'm going over to Universe 2 where the attacker always get's at least 7's.


--Andy

[flexmaster33]

I am so tired of people crying about dice...get over it...the game uses dice...luck is part of the game.

[Cervantes]

There is "luck" as in "Oh, gee, I sure hope I win this roll" or "Darn, I lost a few in a row, shucks".

Then there is "Gee, it's the start of the game, and I just lost 10 attacks in a row... just like the 3 games before it".

And, as someone else pointed out, if you stop and try again next turn, it works out OK. If you keep going, you keep losing. This has happened enough to me that I'm quite sure it's outside of the realm of probability.

I've taken Statistics courses, I understand (a) probability, and (b)human perception (you remember losses more than wins). But I'm quite convinced there is a bias in the dice. This isn't just b!tching about losing a 10 on 2 attack or something. This is "I attack from every territory at the start of the game, split one or two rolls, and lose the rest, and this happens often and has happened more frequently lately".

I've been here a year and a bit now, and this isn't just an unlucky streak (i've had those), this is "something is a bit buggered up somewhere".

[AndyDufresne]

As always, if you are a convinced there is bias you need to:

1. Discredit the current data that says the dice are normal.
2. Prove with your own data that says the dice are not normal.
3. Publish the results and findings (on Conquer Club, or a Scholarly Journal )


--Andy

[Fruitcake]

Ok, here is one example.

There are 91 perms to a 3 dice roll (ignoring the order of course, so 335 is the same as 353 etc).

One assumes from this that over the course of just over 12,000 rolls triple 1 would appear around 130 times.

Why is it my records show it appearing over 5 times that many (669), nearly 1 in 18? When it should be 1 in 91...

Yet triple 2 has appeared but 48 times...

Yes I did start laboriously writing every roll down; as I was certain in my mind something was not right.


Having reviewed this again, the picture actually gets worse.

When analyzing any rolls with at least a double 1 and one other (e.g. 111,112,113,114,115,116) one would assume this should appear 1 in 15 times on average. So the total would be somewhere in the zone of 800 times. However, the total is a staggering 1711 times, double the mathematical average. Drilling down into this, one immediately notices that the higher up the scale (i.e., 114,115,116) you go, the fewer times these numbers appear, the weighting being strongly towards the lower end. 116 actually only appears 98 times. Go figure. I am a mathematician and I truly find these stats beyond comprehension.

Quick update: unsurprisingly this has continued...3 triple ones in the last 42 rolls.

[Fruitcake]

3 more rolls of triple 1 in the last 25....

[wcaclimbing]

heres what we can do to test this:

in a certain "spamalot build game" that i am a part of, we are going to wait until both the attacker and defender have at least 500,000 armies each (500,000 is the number of dice rolls in a file). Then I will reset my dice analyzer, click the auto-attack, and see what i get. it will test the entire dice file at once.

But that wont be for quite a long time, we are all around 16k-17k armies. each set is now worth over 1k each.

[dezzy26]

/\ /\

the length you guys will go to prove or disprove astounds me
ill sit at my computer to await this outcome

hope i dont die of old age lol

[Bean_]

There is "luck" as in "Oh, gee, I sure hope I win this roll" or "Darn, I lost a few in a row, shucks".

Then there is "Gee, it's the start of the game, and I just lost 10 attacks in a row... just like the 3 games before it".

And, as someone else pointed out, if you stop and try again next turn, it works out OK. If you keep going, you keep losing. This has happened enough to me that I'm quite sure it's outside of the realm of probability.

I've taken Statistics courses, I understand (a) probability, and (b)human perception (you remember losses more than wins). But I'm quite convinced there is a bias in the dice. This isn't just b!tching about losing a 10 on 2 attack or something. This is "I attack from every territory at the start of the game, split one or two rolls, and lose the rest, and this happens often and has happened more frequently lately".

I've been here a year and a bit now, and this isn't just an unlucky streak (i've had those), this is "something is a bit buggered up somewhere".

I have noticed the same, and theorize as follows:

"There appear to be some flaws, because the numbers coming from random.org are not simply used once and discarded, but reused and maybe in a faulty manner. I think the effects are (1) a freezing of the probabilities inherent in the original sample (so a small, but semi-permanent deviation from 16-1/6% for each outcome), and (2) an tendency to increase in streakiness (could go either way) on a 13-hour cyclical basis (i.e., the time it takes to pass through the dice file, but I would imagine not very significant until the last 5 minutes of each cycle, which is still 1% of the time)."

http://www.conquerclub.com/forum/viewto ... highlight=

However, despite several attempts at asking for more detail on CC's methodology, I have not received any response. Streakiness is something that the Dice Analyzer can neither prove nor disprove. Even if there is an overall 37% chance that a 3v2 wins, it's not random if at some (rare) times there is a 50% chance of it winning and at other (rare) times there is a 10% chance, which is what may be happening.

[InkL0sed]

http://en.wikipedia.org/wiki/Ganzfeld_experiment