Dice analyzer 1.5.1a, 1.5.1b and 1.5.1c

by **evster** on Wed Aug 30, 2006 7:43 pm

Thanks Cyber!

a) do you have a means of consolidating data?
b) can you move the analyzer to below the chat? On some screens, I already have to scroll down for chat, and it is more cogent game by game than the analyzer.

Thanks
Evan

by **evster** on Wed Aug 30, 2006 10:09 pm

ooh, just realized I have a mixed set...

Hope Patsfan doesn't see this before my next turn!

Ev

by **gavin_sidhu** on Thu Aug 31, 2006 12:59 am

kool. put the link in the first post so everyone can see it.

also dont do what evster is suggesting, that would mean side scrolling which is not good. If u want to put it in a different place, put it under the game chat.

by **evster** on Thu Aug 31, 2006 1:09 am

If you tighten up the graphics, it seems to me there is a lot of free space to the right....

Ev

by **gavin_sidhu** on Thu Aug 31, 2006 2:26 am

evster wrote:If you tighten up the graphics, it seems to me there is a lot of free space to the right....

Ev

I think u must be using the small map for your turns. I use the large one and there is no space at all.

To cyberdaniel: Now what u need to do is either:
a) start a new thread
or:
b) edit ur first post on this thread

to include:
1) the link to install it.
2) a picture showing the expected values u should get after 100 or so throws so i know whether im lucky or not. (right now i have no clue, just know that im not rolling as many 4's as all the other numbers).

by **zip_disk** on Thu Aug 31, 2006 2:52 am

Nappy Bone Apart wrote:I think the more interesting number to me is the 92/45/63. The attacker is coming out ahead in this way more than the raw numbers would show. I appreciate the work you're doing here, I'll be very interested when you get 2-3k rolls for the attacker and 1.5k for the defender.

Well of course the attacker is coming out ahead, the average 3v1 attacker winning percentage should be ~66% and is bloating the overall results. You should just concentrate on one type of combat and the expected outcome for a better comparison.

Btw. Can we get percentages for the lower table as well? Just the attackers win percentage would be enough if it takes up too much space.

by **cyberdaniel** on Thu Aug 31, 2006 1:12 pm

If you really want to get to the chat faster I could make a button to hide/show the stats. Also I could make a single column with the stats.

by **sully800** on Thu Aug 31, 2006 2:01 pm

Pedronicus wrote:
Now all the columns on the defending dice are almost exactly even.

ignore the columns - read the Dice percentages (16.66% is the average we should be expecting) not 18.99% of defending dice are sixes

the more dice are thrown, the smaller the blue bar and less easier to spot a percentage point or two

Yeah and you really shouldn't have the defender throwing less than 13% 5's either. Oh man, the dice are biased against the defender!

I say let a bunch of people use this for a while, then put all the totals into one thread. It should be hard to get 10,000 dice or so like that. If at that point the percentages are off by 2+ percent then it would seem there is a problem IMO

by **PaperPlunger** on Thu Aug 31, 2006 2:42 pm

i love you.

by **gavin_sidhu** on Fri Sep 01, 2006 3:48 am

cyberdaniel wrote:If you really want to get to the chat faster I could make a button to hide/show the stats. Also I could make a single column with the stats.

sounds good. When you update change the name of this thread to version 2 or something so i know when its up and can get the latest. Maybe in your next update you could include the expected values and also change the colours of the attacker and defender dice distribution so people dont get it confused with the key?

by **Pedronicus** on Fri Sep 01, 2006 7:58 am

I've just logged on here after playing a big go on Grand Stratagy.

It's like chalk n cheese.

The dice over there are much better.

by **Phobia** on Fri Sep 01, 2006 10:17 am

i have just tested and so far its great! i have only played a couple of turns, so i cant say much on the results so far

by **onbekende** on Fri Sep 01, 2006 11:43 am

so IE does'nt have it, too bad :cry:

by **cyberdaniel** on Fri Sep 01, 2006 12:11 pm

gavin_sidhu wrote:sounds good. When you update change the name of this thread to version 2 or something so i know when its up and can get the latest.

Ok, I'll do that and post the new version separatelly. People will have the option of choosing between the current one and the new one.

Pedronicus wrote:I've just logged on here after playing a big go on Grand Stratagy.

It's like chalk n cheese.

The dice over there are much better.

The biggest problem with these dice is that they are to streakie. You can have a good day when you kill everything or a bad day when you loose 3v1 8 times in a row. I've noticed that at certain points the defender dice get almost to 16% for every number and times when the attacker dice get to 16% for every number. After that they just keep going into different directions and we end up with lucky streaks or people crying because they lost 8 attacking 1.

by **onbekende** on Fri Sep 01, 2006 12:23 pm

I got a nice one last day, was 14-12, ended 8-1, was funny

by **Pedronicus** on Fri Sep 01, 2006 5:06 pm

14v3 ends up 1v2

Thats not so fucking funny for trying to kill someone with 5 cards.

Website · by **Bad Speler** on Fri Sep 01, 2006 5:08 pm

Its not as bad as I had it one game. Lost 24 vs 9, and then on a seperate attack, desperate for a card, lost 9 vs 1.

by **sully800** on Fri Sep 01, 2006 6:34 pm

cyberdaniel wrote:
gavin_sidhu wrote:sounds good. When you update change the name of this thread to version 2 or something so i know when its up and can get the latest.

Ok, I'll do that and post the new version separatelly. People will have the option of choosing between the current one and the new one.

Pedronicus wrote:I've just logged on here after playing a big go on Grand Stratagy.

It's like chalk n cheese.

The dice over there are much better.

The biggest problem with these dice is that they are to streakie. You can have a good day when you kill everything or a bad day when you loose 3v1 8 times in a row. I've noticed that at certain points the defender dice get almost to 16% for every number and times when the attacker dice get to 16% for every number. After that they just keep going into different directions and we end up with lucky streaks or people crying because they lost 8 attacking 1.

That's exactly how I feel, which I think I already said in this thread. But unfortunately a simple number analysis won't show that because I think in the end the good streaks and bad streaks will even each other out.

Perhaps you could program it to keep track of the longest 3vs1 dice losses and the longest 3vs1 dice wins. If everyone has streaks of that which are way beyond what would be expected it could show the true problem.

by **cyberdaniel** on Fri Sep 01, 2006 7:06 pm

Well, the best thing would be to keep track of all the throws, but I can't store that much info in a cookie. However I'm keeping track of how the stats look every 100 or so throws. So I think that if after a lot of throws we see the streaks still appearing and that the dice still get a lot of a number we can say that the dice aren't random enough.

by **zip_disk** on Fri Sep 01, 2006 8:36 pm

Can it be coded to keep stats for individual games? Premiums might end up with a lot of cookies though. :lol:

by **cyberdaniel** on Sat Sep 02, 2006 12:11 am

It could but you would end up with a lot of cookies, but its a great idea. You could analyze game by game and see how your luck was during a certain game. Do you think its worth making?

by **reverend_kyle** on Sat Sep 02, 2006 2:41 am

by **carmoil** on Sun Sep 03, 2006 6:40 am

Thanks a ton, I've been meaning to start keeping track of my rolls, now I don't have to do any of the work myself. I have about 1500 attacking rolls, and the dice are all within 2% of where they should be, and have been getting closer and closer to even as I keep playing.
I'll post a pic when I have a few more rolls.