1

Onderwerp: spell failures

in speed I have cast 4 spells all failed, is this common with everyone, i know of 2 other players having the same problem

2

Re: spell failures

Not recently, but I used to have similar problems to make me think the random number generator was not working correctly.

I had some very low probability fails.  Like asteroid hitting the earth and destroying all complex life probabilities.

And God Said:
Δ*D = ρf                   Δ*B = 0
ΔxE = -δB/δt            ΔxH = Jf + δD/δt
and THEN there was light.

3

Re: spell failures

% times, then it worked on the sixth, another player has told me 6 out of 10 attempts fail.

4 Laatst bewerkt door Hellikin (2011-09-16 04:52:39)

Re: spell failures

Remember that these events are independent (probabilistically speaking) from each other.  That is, if odds are 50-50, failure in one attempt does not mean that odds shift to greater than 50% entering the next attempt. 

When you cast against an enemy village, if wizard strength, the number of runes in that village spent on wiz towers, are greater than the number of runes spent in your village on wiz towers, then you have a decreased % chance of success.  I assume the relationship is linear, correct?  (i.e. 2 runes in casting town vs 1 rune in castee town = 66% chance of success beyond base chance, or something along those lines)

So, what I'm wondering is if when you are trying to cast a spell "against" your own town, if something in the rune count is not being done properly.  If it's like I described, you should succeed 50% of the time when casting against yourself (i.e. 2 runes vs 2 runes = 50%)

___________________
Edit: Grammar

7 x 13 = 28

Economics today is a race between economists striving to build bigger and better idiot-proof economic or financial models or programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning...

Re: spell failures

I myself have been doubting the random number generator in PHP and I still have a few suspicions (about seeding & server time etc). However, I did some testing a few years earlier and it does seem to be fairly distributed:

http://speed.neuroninteractive.com/test/mtrand

Hellikin schreef:

So, what I'm wondering is if when you are trying to cast a spell "against" your own town, if something in the rune count is not being done properly.  If it's like I described, you should succeed 50% of the time when casting against yourself (i.e. 2 runes vs 2 runes = 50%)

It's a bit more complicated (and also a bit less complicated). Enemy "wizard level" (all runes spent on magic) only gives you a +10 or -10 difficulty bonus, depending on who has more wizards.

The more important ratio is how much runes you've spent on magic in relation to your village size. A good wizard player has 10% of all his runes spent on wizard towers.

6

Re: spell failures

OK, something is not making sense to me.  My outpost casting Rain1 on itself is an 80% probability.  My outpost casting Rain1 on my main is a 90% probability

7 x 13 = 28

Economics today is a race between economists striving to build bigger and better idiot-proof economic or financial models or programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning...

7

Re: spell failures

I had some trouble too. I cast a spell with 90% probability, and it happened already (at least) 4 times that it failed twice in a row. Failing twice has a probability of 1%, failing twice so often (I haven't cast the spell even 20 times I think) is even less. If you'd run a P-test on that (statistics), you'd see that this is beyond normal.

Everyone is entitled to their own opinion. It's just that yours is stupid.

I hate spammer idiots and arrogant players without honour -_-

8 Laatst bewerkt door Hellikin (2011-10-23 15:51:56)

Re: spell failures

I don't think that's an appropriate test for two reasons
1) each cast is an independent event.  Here's an example, the odds of rolling 2 sixes in dice is 11/36, right?   But that is considering that you are rolling together and view them together.  However, the odds of rolling a six the 1st time is 1/6 while the odds of rolling another six with the 2nd roll is also 1/6.  But yes, together it is 11/36.
2) the variable being generated follows the uniform distribution
__________________
Edit: typo

Edit2: To clarify: the odds of success/failure are always the same, regardless of what has happened on the cast before

7 x 13 = 28

Economics today is a race between economists striving to build bigger and better idiot-proof economic or financial models or programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning...

9

Re: spell failures

11/36? I think that is 1/36... the odds of a combined event is the multiplication of the odds of the separate events. 1/6 * 1/6 = 1/36.

Everyone is entitled to their own opinion. It's just that yours is stupid.

I hate spammer idiots and arrogant players without honour -_-

10

Re: spell failures

Chaos schreef:

11/36? I think that is 1/36... the odds of a combined event is the multiplication of the odds of the separate events. 1/6 * 1/6 = 1/36.

Only if you expect the same thing on each throw, or want a certain sequence.  Only then can you look at them as a combined system.

If you are wanting to see the odds of throw a 2... its always 1/6 chance.

And God Said:
Δ*D = ρf                   Δ*B = 0
ΔxE = -δB/δt            ΔxH = Jf + δD/δt
and THEN there was light.

11

Re: spell failures

Well, you cast them in a certain sequence and the odds stay the same... So I don't see why it can't be seen as a combined system...

Everyone is entitled to their own opinion. It's just that yours is stupid.

I hate spammer idiots and arrogant players without honour -_-

12 Laatst bewerkt door Hellikin (2011-10-24 01:04:12)

Re: spell failures

Chaos, because they are independent events (always 1/6), the form is: 1/6+1/6*5/6 = 11/36.  You are not including the 5 out of 6 times where the first roll is not a 6

It is a good game when you can argue probability theory
_______________
Edit: Clarification

7 x 13 = 28

Economics today is a race between economists striving to build bigger and better idiot-proof economic or financial models or programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning...

13

Re: spell failures

I'm just mixed up by how my OP casting on itself has a lower probability than it does when it casts on my main.

7 x 13 = 28

Economics today is a race between economists striving to build bigger and better idiot-proof economic or financial models or programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning...

14 Laatst bewerkt door DoomPope (2011-10-24 04:00:54)

Re: spell failures

Chaos schreef:

Well, you cast them in a certain sequence and the odds stay the same... So I don't see why it can't be seen as a combined system...

If you want to know the probability of one specific sequence you can look at the sequence on the whole.  ie
"What is the chance of getting one 2 in 2 throws?"

would be (1/6)(5/6)+(5/6)(1/6)
Chance of (2, anything else) plus the chance of (anything else, 2)

Meanwhile the question "What is the chance of getting a 2 on the second throw?"
Is simply (1/6) as the first throw has no impact on the outcome of the second throw.

edit:
This is a useful read for calculating probability of multiple attempts
http://en.wikipedia.org/wiki/Binomial_p … l_equation

And God Said:
Δ*D = ρf                   Δ*B = 0
ΔxE = -δB/δt            ΔxH = Jf + δD/δt
and THEN there was light.

15

Re: spell failures

Hmm, I think we don't really understand each other...
I don't want to brag with knowledge or something, but I had statistics class last year and I learned there, that you can solve this kind of problems (determining the odds of 2 fails in a row) with a statistical tree, right? So, the first question is succes or fail, we take the fail branch (10% probability) to the second question (which is the same), and again we go for fail (again 10%). Now, if you want to know the odds of this happening, you have to multiply all odds in every branch you've followed, so that makes 0,1*0,1=0,01=1%.
I don't really see why you should multiply that number again with something else... Independent events just means that the odds of the event doesn't change after one try, or not? If not, statistics in your country is different from here tongue

Everyone is entitled to their own opinion. It's just that yours is stupid.

I hate spammer idiots and arrogant players without honour -_-

16

Re: spell failures

BTW, just happened again... 4500 gems for 1 spell on a net worth of 6300 is pretty much

Everyone is entitled to their own opinion. It's just that yours is stupid.

I hate spammer idiots and arrogant players without honour -_-

17

Re: spell failures

Hellikin schreef:

I'm just mixed up by how my OP casting on itself has a lower probability than it does when it casts on my main.

Your own wizards from the non-casting tower count against you.

Chaos schreef:

Hmm, I think we don't really understand each other...
I don't want to brag with knowledge or something, but I had statistics class last year and I learned there, that you can solve this kind of problems (determining the odds of 2 fails in a row) with a statistical tree, right? So, the first question is succes or fail, we take the fail branch (10% probability) to the second question (which is the same), and again we go for fail (again 10%). Now, if you want to know the odds of this happening, you have to multiply all odds in every branch you've followed, so that makes 0,1*0,1=0,01=1%.
I don't really see why you should multiply that number again with something else... Independent events just means that the odds of the event doesn't change after one try, or not? If not, statistics in your country is different from here tongue

I am going to be overly pedantic and note that this is not statistics, this is probability  wink

I do think for the two failures in a row you are correct.  I always refer to the binomial probability equation for combined probability... because anything with factorials is fun.  Using it the equation reduces to (q)^2|q=.1 = .01 or 1% chance.

And God Said:
Δ*D = ρf                   Δ*B = 0
ΔxE = -δB/δt            ΔxH = Jf + δD/δt
and THEN there was light.

18

Re: spell failures

DoomPope schreef:
Hellikin schreef:

I'm just mixed up by how my OP casting on itself has a lower probability than it does when it casts on my main.

Your own wizards from the non-casting tower count against you.

I can't see how that syncs up with this: http://en.forum.dolumar.com/post/6171/#p6171
In my OP, exactly 10% of runes were invested in wiz towers

I can cast R1 from a weak tower in my main on my main or on my OP with 90% probability.

Now after multiple upgrades to other buildings in my OP, my probability to succeed on a cast of R1 is up to 82% from 80%

7 x 13 = 28

Economics today is a race between economists striving to build bigger and better idiot-proof economic or financial models or programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning...