# Tower/Snake King Event - Hallow's Eve

Page **2** of **3** • 1, **2**, 3

## Re: Tower/Snake King Event - Hallow's Eve

Ive started to press "home" after i press "stop" on the dice and it seems it stopped give 1-1, but not high, so or A-team has something with me after i buycotted and swear them alot in the last period or...i don't know but still lame on them...

And fury, i'm lvl 178 not max level, but if you started to acuse a fair player (26 years old), active and a active forum user of lies or things like that, than please never talk to me in the future or event quote me, act like i don't exist and like my posts are not there please, thanks.

And fury, i'm lvl 178 not max level, but if you started to acuse a fair player (26 years old), active and a active forum user of lies or things like that, than please never talk to me in the future or event quote me, act like i don't exist and like my posts are not there please, thanks.

**MoRandi**- Posts : 339

Join date : 2013-02-15

## Re: Tower/Snake King Event - Hallow's Eve

Y'all realize that there is a truly random element here, right? Said element is your patience; depending on when you press the stop button, you introduce a change in the pseudorandom algorithm. I have personally found that waiting about 3 seconds before pressing stop tends to yield higher sums for me.

~~~~~~~~~~~~~~~~~~~~~~~~

General of Sinnerz (41708568) Member of [S.T.A.R.S.] |

**Max Lvl Party Calc***, Sacrificing Guide & Formation Guide by yours truly

*needs testing with certain teams to ensure quality. also, monster data doesn't update itself without me logging on to manually add monsters, yet, so some of the newest monsters might not be available for use right away. Missing max lvl stats WILL affect final stats, so make sure to pm me with screenshots of it whenever you want to add data.

**Armitaage**- Posts : 1264

Join date : 2012-08-26

Age : 33

## Re: Tower/Snake King Event - Hallow's Eve

Floors 1-75 for me I got a decent amount, usually 5-10. Floors 75-150 my "luck" has disappeared and I get 5 or less almost every roll now, got snake eyes a bunch of times. I find it unlikely to be random if I can get 1-1 5 times but have not yet gotten 6-6 or even 5-6. Not complaining, I always keep my expectations low, just consider my suspicions aroused.

My method has remained unchanged throughout the whole process, tapping constantly to stop as soon as possible to take advantage of bonus time...even when it's not bonus time, I'm impatient :p

My method has remained unchanged throughout the whole process, tapping constantly to stop as soon as possible to take advantage of bonus time...even when it's not bonus time, I'm impatient :p

**echoside**- Posts : 314

Join date : 2013-01-12

## Re: Tower/Snake King Event - Hallow's Eve

Listen - if you managed to read my post, you would see I never accused you of anything. I just fail to see what you expected anyone to think. Do you really expect us to say to ourselves "wow, MoRandi is having unfathomably bad luck, beyond any remotely feasible chance, and everyone else isn't! I guess MoRandi is right and we are all wrong!"MoRandi wrote:Ive started to press "home" after i press "stop" on the dice and it seems it stopped give 1-1, but not high, so or A-team has something with me after i buycotted and swear them alot in the last period or...i don't know but still lame on them...

And fury, i'm lvl 178 not max level, but if you started to acuse a fair player (26 years old), active and a active forum user of lies or things like that, than please never talk to me in the future or event quote me, act like i don't exist and like my posts are not there please, thanks.

You are a fool if that's the response you thought you would get.

If you really did get all 22 rolls without a single combined value higher than 4 (which again - nobody else is even close to that), then that is an extraordinary claim and you shouldn't be surprised if people don't trust you without a shred of evidence.

Again, I'm not saying you are lying, or even saying you exaggerated. I never did. Like you said, A-Team could be out to get you, or maybe the game was bugged, or maybe if you click quickly it has a higher chance of giving low values, or so on. Or maybe you really were that unlucky. Who knows. But don't sound so affronted because you claim something with a chance less than 10^20 happened to you, and nobody else, and someone doesn't trust you 100% right off the bat.

So, basically, chill out. Sometimes, when crazy things happen to you (and only you), people don't believe you. It's cool, that's how the world works : )

**TheFury**- Posts : 235

Join date : 2012-05-27

## Re: Tower/Snake King Event - Hallow's Eve

I haven't gotten 6-6, but I've only gotten one 1-1. I've gotten a couple 6-5s though. I'm only on floor 95 however, so I'll keep an eye if things change in the next 100 floors. I could see them making it harder as you progress (otherwise it'll be easy enough to run out all prizes with a prog and a small set of EP)echoside wrote:Floors 1-75 for me I got a decent amount, usually 5-10. Floors 75-150 my "luck" has disappeared and I get 5 or less almost every roll now, got snake eyes a bunch of times. I find it unlikely to be random if I can get 1-1 5 times but have not yet gotten 6-6 or even 5-6. Not complaining, I always keep my expectations low, just consider my suspicions aroused.

My method has remained unchanged throughout the whole process, tapping constantly to stop as soon as possible to take advantage of bonus time...even when it's not bonus time, I'm impatient :p

**TheFury**- Posts : 235

Join date : 2012-05-27

## Re: Tower/Snake King Event - Hallow's Eve

Well I started thinking about it, 1 roll every 5 floors, a bonus roll every 20, at floor 150 I've done 37 rolls and got approx. 220 candies, so it seems my later bad luck was to balance out my previous better luck? Cause that would average out to pretty much 6 for every roll. I must just remember the bad rolls more.

**echoside**- Posts : 314

Join date : 2013-01-12

## Re: Tower/Snake King Event - Hallow's Eve

Well, the average SHOULD be 7. So you join the group of the majority of people who are getting 0-1 candy fewer than they should be. I'll keep an eye on my clanmates to see if maybe it's just luck of the draw, of if there really is a slight skew downward.echoside wrote:Well I started thinking about it, 1 roll every 5 floors, a bonus roll every 20, at floor 150 I've done 37 rolls and got approx. 220 candies, so it seems my later bad luck was to balance out my previous better luck? Cause that would average out to pretty much 6 for every roll. I must just remember the bad rolls more.

As Armitaage pointed out, it could also just be that A-Team punishes people who click too quickly (or something) on the stop button, and that's what is skewing results.

**TheFury**- Posts : 235

Join date : 2012-05-27

## Re: Tower/Snake King Event - Hallow's Eve

Update: having finished 155 floors and 38 rolls, I've had exactly 250 candies. That averages to 6.57 per roll, which is close to where it should be if it was not rigged.

~~~~~~~~~~~~~~~~~~~~~~~~

General of Sinnerz (41708568) Member of [S.T.A.R.S.] |

**Max Lvl Party Calc***, Sacrificing Guide & Formation Guide by yours truly

*needs testing with certain teams to ensure quality. also, monster data doesn't update itself without me logging on to manually add monsters, yet, so some of the newest monsters might not be available for use right away. Missing max lvl stats WILL affect final stats, so make sure to pm me with screenshots of it whenever you want to add data.

**Armitaage**- Posts : 1264

Join date : 2012-08-26

Age : 33

## Re: Tower/Snake King Event - Hallow's Eve

i'm at floor 217 now. at floor 200 I had 327 pieces of candy. Got rss on 9th or 10th exchange

~~~~~~~~~~~~~~~~~~~~~~~~

ID: 1811262341

**Rinoa**- Posts : 33

Join date : 2013-03-29

Age : 38

Location : Bishkek, Kyrgyzstan

## Re: Tower/Snake King Event - Hallow's Eve

Fury don't react like that, but that's how it sounded, when i am pissed of and angry about that unluck or wtf it is and after i read what you said ofc it made me more frustrated.

I see people here-there got rss in 2-9 tryes, got nice dice rolls how its happening that only me i get nothing, on 4 accounts i did alot rolls and got only silver golerms, some SOWs and a few gold golems, a real life friend of mine was at my place, he got in 3rd try so i got more pissed off...

To complete all this, fought today 2 divinityes and got 2 Ex stones, and with today fights i did more than 20 in total only in weekends...

To be honest maybe i aswell haven't belive 100% if i were you or any other player, but now, after what i happening to me, i would belive anyone else.

1 thing is clear, i don't wish you to get to belive me by happening all this to you too...

I am in general an unlucky person in games, and after the last event when i tought my bad luck finally have gone, now it happens all these things, at an exagerated rate, to compensate the missing bad luck...or a-team agains me, or what the hell, i don't know

I'm a trusted player and i don't want to lose that because some players doesn't belive my words, if i can turn the time i wouldn't have post that unluck or what it is, i would have keep it for me and it was better...

I see people here-there got rss in 2-9 tryes, got nice dice rolls how its happening that only me i get nothing, on 4 accounts i did alot rolls and got only silver golerms, some SOWs and a few gold golems, a real life friend of mine was at my place, he got in 3rd try so i got more pissed off...

To complete all this, fought today 2 divinityes and got 2 Ex stones, and with today fights i did more than 20 in total only in weekends...

To be honest maybe i aswell haven't belive 100% if i were you or any other player, but now, after what i happening to me, i would belive anyone else.

1 thing is clear, i don't wish you to get to belive me by happening all this to you too...

I am in general an unlucky person in games, and after the last event when i tought my bad luck finally have gone, now it happens all these things, at an exagerated rate, to compensate the missing bad luck...or a-team agains me, or what the hell, i don't know

I'm a trusted player and i don't want to lose that because some players doesn't belive my words, if i can turn the time i wouldn't have post that unluck or what it is, i would have keep it for me and it was better...

**MoRandi**- Posts : 339

Join date : 2013-02-15

## Re: Tower/Snake King Event - Hallow's Eve

Well if it makes you feel any better MoRandi I've done 71 candy trades on my main account and still have not gotten the rss or either of the AA summons, I'm on floor 420-something. There are 29 prizes left and I'm pretty sure I know what the very last 3 will be.

**echoside**- Posts : 314

Join date : 2013-01-12

## Re: Tower/Snake King Event - Hallow's Eve

Using the dice rolls posted by Armitaage, Echo and Rinoa, we're already 3 standard deviations from 7, without even having to add in the rolls from the 4 accounts I'm tracking, which are all performing similarly. Anyone not convinced yet?

**ryesteve**- Posts : 1418

Join date : 2012-05-25

## Re: Tower/Snake King Event - Hallow's Eve

Well I was wrong, took me 81 tries but I got a random skill spirit, 82 was an AA summon (dru)

I'm on floor 477, did 82 exchanges and have 1 single candy left over, so if my math is correct that was 118 rolls resulting in 821 candies which would be an average of 6.95, average has gone up 1 whole point.

I'm on floor 477, did 82 exchanges and have 1 single candy left over, so if my math is correct that was 118 rolls resulting in 821 candies which would be an average of 6.95, average has gone up 1 whole point.

**echoside**- Posts : 314

Join date : 2013-01-12

## Re: Tower/Snake King Event - Hallow's Eve

Steve, I'm not sure how you're getting 3 standard deviations. Values within ± 1sd should account for about 68%, within ± 2 sd for about 95%, and within ± 3 sd for about 99.7%, in a normal distribution. Our values of about 6.5 easily fall within 1 sd of the expected mean, 7. We would all need to be averaging less than poor Mo to be within 3 sd. 68% encompasses from just below 5 to just above 9, 95% is from just below 3 to just above 11, and 99.7% encompasses all possible sums. Run a sd of the possible values and you will see that the sd is about 2.44, which puts our averages well within 1 sd(except for poor Mo; kinda sucks to be on the wrong end of improbability, I can only imagine)

~~~~~~~~~~~~~~~~~~~~~~~~

General of Sinnerz (41708568) Member of [S.T.A.R.S.] |

**Max Lvl Party Calc***, Sacrificing Guide & Formation Guide by yours truly

*needs testing with certain teams to ensure quality. also, monster data doesn't update itself without me logging on to manually add monsters, yet, so some of the newest monsters might not be available for use right away. Missing max lvl stats WILL affect final stats, so make sure to pm me with screenshots of it whenever you want to add data.

**Armitaage**- Posts : 1264

Join date : 2012-08-26

Age : 33

## Re: Tower/Snake King Event - Hallow's Eve

I can see where you are coming from, and yeah I worded my response more antagonistically than I should have (and I apologize for that), but it's like two people are talking about a government plan and how it increased or decreased their taxes by a few hundred bucks, then a third person comes in and says the government kicked open their door, robbed them of all their money, smeared mud in their eyes, then kicked them in the shins and ran away . Considering how often people over-exaggerate things on the internet (myself included), that just seems like a much more likely explanation.MoRandi wrote:Fury don't react like that, but that's how it sounded, when i am pissed of and angry about that unluck or wtf it is and after i read what you said ofc it made me more frustrated.

I see people here-there got rss in 2-9 tryes, got nice dice rolls how its happening that only me i get nothing, on 4 accounts i did alot rolls and got only silver golerms, some SOWs and a few gold golems, a real life friend of mine was at my place, he got in 3rd try so i got more pissed off...

To complete all this, fought today 2 divinityes and got 2 Ex stones, and with today fights i did more than 20 in total only in weekends...

To be honest maybe i aswell haven't belive 100% if i were you or any other player, but now, after what i happening to me, i would belive anyone else.

1 thing is clear, i don't wish you to get to belive me by happening all this to you too...

I am in general an unlucky person in games, and after the last event when i tought my bad luck finally have gone, now it happens all these things, at an exagerated rate, to compensate the missing bad luck...or a-team agains me, or what the hell, i don't know

I'm a trusted player and i don't want to lose that because some players doesn't belive my words, if i can turn the time i wouldn't have post that unluck or what it is, i would have keep it for me and it was better...

That being said, you do post commonly here and the last thing the dwindling forum community needs is infighting. Plus, you are a pretty reasonable guy all things considered, so I'll trust you on this.

As more reference, got two more bits of information from clan members once they get to 20 rolls. 138 and 132 candies each. So something screwy is going on with just you Mo. Just bad luck can't come close to explaining your low numbers - the chances of that happening are far, far, far below one in a trillion trillion - so must be a bug with your account or something. Or someone in A-Team has it out for you haha. Best of luck figuring it out, and I hope it doesn't continue.

Last edited by TheFury on Sat Oct 19, 2013 4:46 pm; edited 1 time in total

**TheFury**- Posts : 235

Join date : 2012-05-27

## Re: Tower/Snake King Event - Hallow's Eve

Would you mind updating us when you get all 100 exchanges? I'm curious if the rewards just stop, or if it resets, etc etc.echoside wrote:Well I was wrong, took me 81 tries but I got a random skill spirit, 82 was an AA summon (dru)

I'm on floor 477, did 82 exchanges and have 1 single candy left over, so if my math is correct that was 118 rolls resulting in 821 candies which would be an average of 6.95, average has gone up 1 whole point.

**TheFury**- Posts : 235

Join date : 2012-05-27

## Re: Tower/Snake King Event - Hallow's Eve

I got all but the last two, a pack of 5 yellow stones and a plat golem, there is a reset prizes button at the bottom.

**echoside**- Posts : 314

Join date : 2013-01-12

## Re: Tower/Snake King Event - Hallow's Eve

Armitaage, you're forgetting to divide the 2.44 by the square root of the sample size

**ryesteve**- Posts : 1418

Join date : 2012-05-25

## Re: Tower/Snake King Event - Hallow's Eve

Data set:

{2,3,3,4,4,4,5,5,5,5,6,6,6,6,6,7,7,7,7,7,7,8,8,8,8,8,9,9,9,9,10,10,10,11,11,12}

Total number of items: 36

Sum: 252

Mean: 7

Sd formula:

The differences between the individual numbers and the mean range from -5 to 5 as follows:

-5×1

-4×2

-3×3

-2×4

-1×5

0×6

1×5

2×4

3×3

4×2

5×1

Therefore their squares are:

25×2

16×4

9×6

4×8

1×10

0×6

The sum of that is 210. Divided by the size of the set, 36, we get 5.833333333. Square root of that is 2.415229457, which is the standard deviation. Where are you getting dividing by the square root of the size again?

{2,3,3,4,4,4,5,5,5,5,6,6,6,6,6,7,7,7,7,7,7,8,8,8,8,8,9,9,9,9,10,10,10,11,11,12}

Total number of items: 36

Sum: 252

Mean: 7

Sd formula:

The differences between the individual numbers and the mean range from -5 to 5 as follows:

-5×1

-4×2

-3×3

-2×4

-1×5

0×6

1×5

2×4

3×3

4×2

5×1

Therefore their squares are:

25×2

16×4

9×6

4×8

1×10

0×6

The sum of that is 210. Divided by the size of the set, 36, we get 5.833333333. Square root of that is 2.415229457, which is the standard deviation. Where are you getting dividing by the square root of the size again?

~~~~~~~~~~~~~~~~~~~~~~~~

General of Sinnerz (41708568) Member of [S.T.A.R.S.] |

**Max Lvl Party Calc***, Sacrificing Guide & Formation Guide by yours truly

*needs testing with certain teams to ensure quality. also, monster data doesn't update itself without me logging on to manually add monsters, yet, so some of the newest monsters might not be available for use right away. Missing max lvl stats WILL affect final stats, so make sure to pm me with screenshots of it whenever you want to add data.

**Armitaage**- Posts : 1264

Join date : 2012-08-26

Age : 33

## Re: Tower/Snake King Event - Hallow's Eve

Sample size is a basic component of any test of significance. But I'm pretty sure you already must know that.

Here, go to town: http://en.wikipedia.org/wiki/Z-test

Here, go to town: http://en.wikipedia.org/wiki/Z-test

**ryesteve**- Posts : 1418

Join date : 2012-05-25

## Re: Tower/Snake King Event - Hallow's Eve

I fail to see how a z-test is appropriate here. Why not show the work you have done, to illustrate what you're trying to show? Oh, and do find a source other than Wikipedia. This might come as a surprise, but college professors don't accept Wikipedia articles as an appropriate reference.

Additionally, after reading up on the subject, what you're calculating by dividing by the square root of the sample size is the

Furthermore, the numbers that I crunched in the last post are of the fair dice, therefore a z-test of those numbers will only confirm that it is indeed a fair dice. What you really need to do a z-test of is each individual player's rolls. Unfortunately, here you have a problem: you only have total number of rolls and the sum of those rolls, but you don't have the actual rolls themselves. That said, you have no way of computing the variance of each player, let alone their standard deviation or standard error. What is under scrutiny is the game's die rolls, not the control dice rolls.

Additionally, after reading up on the subject, what you're calculating by dividing by the square root of the sample size is the

**, not the standard deviation. Standard error is equal to the square root of the variance divided by the sample size. The variance is the square of the standard deviation.***standard error*Furthermore, the numbers that I crunched in the last post are of the fair dice, therefore a z-test of those numbers will only confirm that it is indeed a fair dice. What you really need to do a z-test of is each individual player's rolls. Unfortunately, here you have a problem: you only have total number of rolls and the sum of those rolls, but you don't have the actual rolls themselves. That said, you have no way of computing the variance of each player, let alone their standard deviation or standard error. What is under scrutiny is the game's die rolls, not the control dice rolls.

~~~~~~~~~~~~~~~~~~~~~~~~

General of Sinnerz (41708568) Member of [S.T.A.R.S.] |

**Max Lvl Party Calc***, Sacrificing Guide & Formation Guide by yours truly

*needs testing with certain teams to ensure quality. also, monster data doesn't update itself without me logging on to manually add monsters, yet, so some of the newest monsters might not be available for use right away. Missing max lvl stats WILL affect final stats, so make sure to pm me with screenshots of it whenever you want to add data.

**Armitaage**- Posts : 1264

Join date : 2012-08-26

Age : 33

## Re: Tower/Snake King Event - Hallow's Eve

A z-test is appropriate because we know the population mean and variance. and we want to know if the sample of observations differs from that population.

I wasn't citing Wikipedia as a "reference", I was citing it as an article that was explaining it at a level I figured you would understand. I guess I was wrong. If I tell you that the variance you computed needs to be divided by the square root of the sample size, and you tell me to "show my work", you're operating at a level far below what I gave you credit for. Or you're just being a troll.

You don't need the rolls to compute the variance, because it's the variance of two fair dice that's the appropriate parameter, and we already know what that is. If you want to take the position that the sample variance is something different, then you've already answered the question that these aren't two fair dice.

So seriously, you're telling me that you don't see how sample size factors in here? So if I'm averaging 6.5, in your world, it doesn't matter if came from 2 rolls or 2 million, it'd still be really, really close to the mean for you?

I feel like we're at the point now where I could tell the sky is blue, and you'd push back because I'm the one saying it. I'm done taking your troll bait. Ask TheFury to explain it to you further.

I wasn't citing Wikipedia as a "reference", I was citing it as an article that was explaining it at a level I figured you would understand. I guess I was wrong. If I tell you that the variance you computed needs to be divided by the square root of the sample size, and you tell me to "show my work", you're operating at a level far below what I gave you credit for. Or you're just being a troll.

You don't need the rolls to compute the variance, because it's the variance of two fair dice that's the appropriate parameter, and we already know what that is. If you want to take the position that the sample variance is something different, then you've already answered the question that these aren't two fair dice.

So seriously, you're telling me that you don't see how sample size factors in here? So if I'm averaging 6.5, in your world, it doesn't matter if came from 2 rolls or 2 million, it'd still be really, really close to the mean for you?

I feel like we're at the point now where I could tell the sky is blue, and you'd push back because I'm the one saying it. I'm done taking your troll bait. Ask TheFury to explain it to you further.

**ryesteve**- Posts : 1418

Join date : 2012-05-25

## Re: Tower/Snake King Event - Hallow's Eve

In other words, I'm not being the troll. Over the last 3 posts you have made, you have shown that you really don't know what you are talking about.

First, showing your work. In order to conclusively prove anything, you need to show to the people who doubt you exactly how you arrive at your conclusion, so that they can review your work and draw their own conclusion. Otherwise, as far as they are concerned, you've just made up a result that suits you and thrown a few random scientific words in the sentence to scare them away from trying to scrutinize your work. In other words, your claim becomes little more than hearsay. If you actually didn't make it up, you need to show how you got there.

Secondly, you seem to know very little about the scientific method in general and the z-test in particular. I have gone through the painstaking research to get a clear picture of the z-test from sources other than Wikipedia(which has "citation needed" markers all over the statistical math articles, including the one for the z-test. It's why I'm discounting it; any article that is missing citations on a subject that is taught in AP courses in high school is one I really don't trust in the least). The actual z-test has several variations, but all of them require the same information:

The population average

The population standard deviation

The sample average

The sample size(this much you had correct, at one point or another)

A confidence level(seems to be not stated in any of your posts, yet it is a very important point)

A z-score table(kindly provided by various universities, if you do a little research)

Since we're checking to see if the sample average is significantly lower than the population average, I chose to perform a 1 tailed test. Here's the steps that I took:

For each of the three samples (220/37, 250/38, and 327/50), I subtracted the population average from the sample average getting:

-1.055, -0.422, -0.460

Then I divided the population standard deviation by the square root of each sample's size, getting:

0.397, 0.391, 0.341

Next, I divided the first set of numbers by the second set, getting these z-scores:

-2.6574, -1.0792, -1.3489

Which I then checked against the z-score table linked to by the ehow article on z-tests. When my numbers had a higher precision than the table, I rounded down. These are the p-values for each sample:

0.0040, 0.1423, 0.0901

Now, depending on what confidence level was initially chosen, we get different results from the test. At a confidence level of 99.6%, not one of the averages are considered to be significantly lower than the population's average. At 95%, the 220/37 sample's average is considered to be significantly lower than the population's. At 90%, both the 220/37 and the 327/50 are considered to be significantly lower. It is not until 85% confidence level that all the samples are significantly lower than the population. It is your omittance of the confidence level and your juggling terms randomly(variance is not the same as standard deviation and standard deviation is certainly not the same as standard error or z-score) that makes me think that you know even less than me on the subject of z-tests. Not only that, but the z-test doesn't tell you how many standard deviations away from the true mean a sample's mean is. It just tells you how significantly close the two averages are.

Honestly, before you try to use math to argue your point, make sure that you're doing it right. And don't call people names just because they ask you to show your work, it just makes YOU look bad.

References:

http://www.ehow.com/how_8476911_calculate-ztest.html

http://lilt.ilstu.edu/dasacke/eco148/ztable.htm

First, showing your work. In order to conclusively prove anything, you need to show to the people who doubt you exactly how you arrive at your conclusion, so that they can review your work and draw their own conclusion. Otherwise, as far as they are concerned, you've just made up a result that suits you and thrown a few random scientific words in the sentence to scare them away from trying to scrutinize your work. In other words, your claim becomes little more than hearsay. If you actually didn't make it up, you need to show how you got there.

Secondly, you seem to know very little about the scientific method in general and the z-test in particular. I have gone through the painstaking research to get a clear picture of the z-test from sources other than Wikipedia(which has "citation needed" markers all over the statistical math articles, including the one for the z-test. It's why I'm discounting it; any article that is missing citations on a subject that is taught in AP courses in high school is one I really don't trust in the least). The actual z-test has several variations, but all of them require the same information:

The population average

The population standard deviation

The sample average

The sample size(this much you had correct, at one point or another)

A confidence level(seems to be not stated in any of your posts, yet it is a very important point)

A z-score table(kindly provided by various universities, if you do a little research)

Since we're checking to see if the sample average is significantly lower than the population average, I chose to perform a 1 tailed test. Here's the steps that I took:

For each of the three samples (220/37, 250/38, and 327/50), I subtracted the population average from the sample average getting:

-1.055, -0.422, -0.460

Then I divided the population standard deviation by the square root of each sample's size, getting:

0.397, 0.391, 0.341

Next, I divided the first set of numbers by the second set, getting these z-scores:

-2.6574, -1.0792, -1.3489

Which I then checked against the z-score table linked to by the ehow article on z-tests. When my numbers had a higher precision than the table, I rounded down. These are the p-values for each sample:

0.0040, 0.1423, 0.0901

Now, depending on what confidence level was initially chosen, we get different results from the test. At a confidence level of 99.6%, not one of the averages are considered to be significantly lower than the population's average. At 95%, the 220/37 sample's average is considered to be significantly lower than the population's. At 90%, both the 220/37 and the 327/50 are considered to be significantly lower. It is not until 85% confidence level that all the samples are significantly lower than the population. It is your omittance of the confidence level and your juggling terms randomly(variance is not the same as standard deviation and standard deviation is certainly not the same as standard error or z-score) that makes me think that you know even less than me on the subject of z-tests. Not only that, but the z-test doesn't tell you how many standard deviations away from the true mean a sample's mean is. It just tells you how significantly close the two averages are.

Honestly, before you try to use math to argue your point, make sure that you're doing it right. And don't call people names just because they ask you to show your work, it just makes YOU look bad.

References:

http://www.ehow.com/how_8476911_calculate-ztest.html

http://lilt.ilstu.edu/dasacke/eco148/ztable.htm

~~~~~~~~~~~~~~~~~~~~~~~~

General of Sinnerz (41708568) Member of [S.T.A.R.S.] |

**Max Lvl Party Calc***, Sacrificing Guide & Formation Guide by yours truly

*needs testing with certain teams to ensure quality. also, monster data doesn't update itself without me logging on to manually add monsters, yet, so some of the newest monsters might not be available for use right away. Missing max lvl stats WILL affect final stats, so make sure to pm me with screenshots of it whenever you want to add data.

**Armitaage**- Posts : 1264

Join date : 2012-08-26

Age : 33

## Re: Tower/Snake King Event - Hallow's Eve

Hey everyone,

I'm new to the forum but have been playing DS for about 100 days, I'm level 81. Cant say that I have kept an exact count so cant provide the mean of my dice rolls, but I think it would be around 8 (I have been rather lucky most the time). Also I have already gained 2 rss from the candy corner (as soon as I recieve it I reset the prizes to try and get another.) I'm on floor 225 at the moment so I think I have been very lucky to get the prizes I have so far. Just wanted to throw that in there as I do believe that A-Team have made this random, certain people have obviously been extremely unlucky but there are extremes on both ends of the scale. I am going for as many rss as possible by the end of the event I wont argue maths with you guys, I consider myself good with numbers, but reading some of that is making my brain melt!

I'm new to the forum but have been playing DS for about 100 days, I'm level 81. Cant say that I have kept an exact count so cant provide the mean of my dice rolls, but I think it would be around 8 (I have been rather lucky most the time). Also I have already gained 2 rss from the candy corner (as soon as I recieve it I reset the prizes to try and get another.) I'm on floor 225 at the moment so I think I have been very lucky to get the prizes I have so far. Just wanted to throw that in there as I do believe that A-Team have made this random, certain people have obviously been extremely unlucky but there are extremes on both ends of the scale. I am going for as many rss as possible by the end of the event I wont argue maths with you guys, I consider myself good with numbers, but reading some of that is making my brain melt!

**Psycom69**- Posts : 1

Join date : 2013-10-20

Age : 31

## Re: Tower/Snake King Event - Hallow's Eve

All I'm going to say is I got the RSS on my second try cashing in my candies, used it on my AAA+ and got slumbering breath.

I got the AA summon on my 4th try, used it and got a Vlad.

Why are we concerned about the randomness of the dice when we should be concerned about the true randomness of summoning anything other than a Formless Nyar with our AA summons, or imparting any of our hard-won monsters with a skill other than Crit Rate Up from these RSS?

I dig all the math stuff, tho. Don't understand it, but I appreciate the analyses

I got the AA summon on my 4th try, used it and got a Vlad.

Why are we concerned about the randomness of the dice when we should be concerned about the true randomness of summoning anything other than a Formless Nyar with our AA summons, or imparting any of our hard-won monsters with a skill other than Crit Rate Up from these RSS?

I dig all the math stuff, tho. Don't understand it, but I appreciate the analyses

~~~~~~~~~~~~~~~~~~~~~~~~

**Ologolos**- Posts : 565

Join date : 2012-12-03

Age : 40

Location : Pittsburgh, PA

Page **2** of **3** • 1, **2**, 3

Page

**2**of**3****Permissions in this forum:**

**cannot**reply to topics in this forum