attack and defense bonuses dont really work!

[Santa367]

Surely im not the 1st who noticed that, but i just wanna make it clear for all people.

Here is some maths about 5% attack and defense bonuses. 5% bonus doesnt mean 5% bigger winning chance.
For example demons or vampires attack any non def bonus race : Weve got 8v8 fight so that there are
(48-6+1)*(48-6+1)=1849 possible results.
Now lets count the results which would make our attack bonus work: rolling 20-39 while enemy rolls the same , so there 20 possible results leading us to the win. Also rolling 40-48 while enemy rolls same or +1, so there are 18 results more. Summing up weve got 38 cases of 1849 in which bonus works. So it increase the winning chance by only
38/1849=0,0205%.

Winning chances increase for the def bonus is the same.

As u can see it would work even worse if fighting with 4-7 and wouldnt work at all with 3-1.

Imao all races with those bonuses suck and Pumpkins is the best race then.

I hope there arent any mistakes at my maths and english=)

Looking forward to ur comments;)

[Jokern]

omg ::O

[Darkas]

Wrong math are wrong.

[Santa367]

Wrong math are wrong.

what exactly? correct it then!

[Darkas]

Let's take only two dice as example.

You can roll a sum between 2 and 12, but the probability of each sum are not equal.

You have 6*6=36 possible rolls.

To make 2, you have only one option: 1 & 1 -> 1/36
To make 3, you have 1 & 2 and 2 & 1 -> 2/36
To make 4, 1 & 3, 2 & 2, 3 & 1 -> 3/36
..
To make 7, 6/36
To make 8, 5/36.
...
To make 12, 1/36.

--

You have to weight each output with its probability, you arbitrary, and wrongly, put the same probability at each sum (rolling a 30 is very, very more common than rolling an 8 or a 48).

[Santa367]

yeah, thats right, i was false, too lazy to make the math now: could u maybe? do u have a program that could calculate it?
But in any case im sure that the winning probability increase will stay much less then 5% just cause of the statement that it works only if enemy rolls the same or eventually +1

[techgump]

oh dear....

Santa, read these posts:
viewtopic.php?f=8&t=966
and
viewtopic.php?f=4&t=1107

[Darkas]

But in any case im sure that the winning probability increase will stay much less then 5% just cause of the statement that it works only if enemy rolls the same or eventually +1.

The end of your sentence is right, but the beginning is wrong :P

I actually did a lil' program which compute all combinations (-> exact numbers, not simulations).

For a 8v8, you goes from 47.1% to 52.9% with the attack bonus (demons, vampires). It is of course the higher modification with 7v7, but... 8v8 are quite common...


Note: Usually, the defender win on equality, therefore the 47.1 is realistic. With the bonus, and as the mean is 30, it's likely that it goes up at least to 50%, and even a little more because you goes from 8 to 48, and the bonus is worth from 20 to 48, and makes you win on +1 from 40 to 48(47 assuming +1).

[Santa367]

oh dear....

Santa, read these posts:
viewtopic.php?f=8&t=966
and
viewtopic.php?f=4&t=1107

done

no one of them contains any math to my topic: i was talking neighter about normal winning chances nor about winning chances using boosts.
I just want to calculate winning chance increase using a race with attack bonus

[Darkas]

See my previous post.

The exact (epsilon = .0001%, absolute) modifier (increase) for 8v8 is 5.7463%.

And, guess what, that's relatively to your chance of victory more than 10% (~12.2%)!