Shakes & Fidget - The Game - Supportboard  

Go Back   Shakes & Fidget - The Game - Supportboard > Board Archives > Archive Game & Support

Notices

 
 
Thread Tools Display Modes
  #41  
Old 10-23-2011, 08:29 AM
Bullbound Bullbound is offline
Supreme Adventurer
 
Join Date: Apr 2010
Posts: 3,208
Default

I'm saying you can not look at the number of players, but have to look at the number of accounts. Each new account is another character that the events could occur to. If you want to see if something is significant, you can not look at the number of player. Each person is allowed another account on every server. I have 7 accounts, for example. I have one on US s1, US s2, US s3, US s4, US s5, US s6, and US s7. If you count that as only one player, then you are ignoring the fact that I am contributing to the ratio more than once.
__________________
[SIGPIC][/SIGPIC]
Please click "Thanks" if you found my post helpful
  #42  
Old 10-23-2011, 11:15 AM
Darstard Darstard is offline
Adventurer
 
Join Date: May 2010
Location: Canada
Posts: 393
Default

I'm not sure what you guys are debating any more. Yes a run of 35 epic-less dungeons is very unlikely. So is winning the lottery or getting hit by lighting, yet worldwide those things happen routinely.
Cregan, your math proved that it CAN happen and it did. Is the game broken? No. Does it stink that it happened to a player in our community. Sure it does.
__________________
[SIGPIC][/SIGPIC]
  #43  
Old 10-23-2011, 12:27 PM
Cregan Cregan is offline
Mushroom Collector
 
Join Date: Oct 2011
Posts: 37
Default

Dastard:

What I showed is that given the chance of 25% it's so very unlikely that a streak of 35 (or better yet 89) happens that there is mathematical grounds to doubt the given 25%. Nothing less, nothing more.

Yes, it CAN happen. But me getting a toothpick in my left eye while riding an icebear in africa while humming the Mythbuster theme CAN happen as well.

Last edited by Cregan; 10-23-2011 at 12:47 PM.
  #44  
Old 10-24-2011, 06:05 AM
Darstard Darstard is offline
Adventurer
 
Join Date: May 2010
Location: Canada
Posts: 393
Default

Assuming that Derpa is 100% accurate, I would say you're part way there, but your sample group is quite skewed. You're only going to get the outliers complaining about epic-less runs, so you can't really be sure what the actual playing population is.
Also, let's go back and review your input data: Derpa's 35 epic-less dungeon run.

10/7 - L37 toon completed 17 dungeons. D17 in order of level (the most likely way he's going to complete them) is an L42.
10/20 - Lxx toon has now completed 38 dungeons? (vs an L84?)
10/21 - Lxx toon has now completed 39 dungeons (vs an L86)

Running 21 dungeons in 2 weeks seems so unlikely to me that I think he's switched from reporting his dungeon count to his level. Meaning he's only cleared 18 dungeons and received 4 epics for it. Even if I'm mistaken here and he has actually cleared 39 dungeons, without supporting data, we can't be sure that he's talking about found vs kept. I've had long runs where I haven't KEPT an epic as the drops were not as good as ones I already had. Humans readily change the language around those types of events from an accurate statement of the occurance to something that sounds more like "I haven't got one in so long...", which can actually change their own memory of what's occurred.

I'm NOT saying that Derpa is lying, only that as a human eye witness he is quite fallible, as well as being far too small a sample size to base any true statistical analysis on. So while there's nothing wrong with your math, the conclusion you've drawn from your work (doubting the stated 25% chance) is not really supported. At best you could draw a conclusion that you might want to repeat the analysis with more data that is verifiable.

<-- A smile and a wink to ensure that folks know I'm enjoying the challenge of this discussion in a fun way. I'm don't think my tone above communicates that properly.

__________________
[SIGPIC][/SIGPIC]

Last edited by Darstard; 10-24-2011 at 09:09 AM. Reason: Added smiles, wink and smile/wink text.
The Following User Says Thank You to Darstard For This Useful Post:
DominicS (12-30-2011)
  #45  
Old 10-24-2011, 12:01 PM
Derpa Derpa is offline
Sheep
 
Join Date: Oct 2011
Posts: 51
Default

Staff of the shrunken head
Eerie mummy grasp of Osiris
Gloves of something or other
Staff of the Dragonlord
Treasure map of empty promises

I think i may have possibly had 1 other but certainly not more than that, count now stands at 41 dungeons.
  #46  
Old 10-24-2011, 01:06 PM
Darstard Darstard is offline
Adventurer
 
Join Date: May 2010
Location: Canada
Posts: 393
Default

41 dungeons? That's 24 dungeons in 16 days? And you weren't saving them before? What level are you now? You were according to your earlier post, L38 on October 8.
__________________
[SIGPIC][/SIGPIC]
  #47  
Old 10-24-2011, 01:32 PM
Charista Charista is offline
Adventurer
 
Join Date: Jun 2010
Posts: 448
Default

Quote:
Originally Posted by Cregan View Post
Dastard:

What I showed is that given the chance of 25% it's so very unlikely that a streak of 35 (or better yet 89) happens that there is mathematical grounds to doubt the given 25%. Nothing less, nothing more.

Yes, it CAN happen. But me getting a toothpick in my left eye while riding an icebear in africa while humming the Mythbuster theme CAN happen as well.
And my point is that, no, it isn't.

If it can happen, it will. And when it does, they'll come here and complain. The tens of thousands of people to which it does not happen will not represent themselves here. Consequently, if a 1/25000 chance happens in a sample space of 100000, you'll find four people complaining the odds are wrong, but the apparent sample space of the forum is only 1000, giving the appearance of a 1/250 chance, when it really was 1/25000. Complaints to this forum are not evidence of statistical failure.

We are not able to determine if the odds of epics dropping are wrong from anecdotal evidence. There *WILL* be someone like me with a large drought, and you *WILL* find that person even if the chances are 25%. Like I said, if it can happen, it will happen to someone. Finding that someone does not cast doubt on the chance of something rare happening.

Last edited by Charista; 10-24-2011 at 01:34 PM.
  #48  
Old 10-24-2011, 02:23 PM
Cregan Cregan is offline
Mushroom Collector
 
Join Date: Oct 2011
Posts: 37
Default

Dastard and Charista, you make valid points, on which I will have to do some thinking I'm loving these mental excersices (spelling?), so this discussion is very welcome

I'm going to present a different setting, in the hope I can convey why I think sample-size is irrelevant, and in the hope I can understand where my or your reasoning is off

You are playing a simple game with friends. Each player has a six-sided die of his own. The one to roll a 6 gets to eat some really, really delicious chips. The rest is stuck with some really bland cookies, that is, until they roll a 6 themselves.

After a little (say 30 minutes and many die-rolls later) while you notice a friend getting to eat a lot of the chips, so you decide to keep track of the number of sixes he rolls.

Of the 80 subsequent rolls he makes, 35 are sixes! Is this a coincidence? Luckily one of your friends knows a bit of math and you decide to calculate if the die is loaded or not. You agree beforehand that rolls with less than 1% chance are so unlikely to happen that, in that case, the die must be loaded.

The chance of rolling 35 or more sixes in 80 throws is 0,000001 % for a normal die.

Mathematically seen, this is nigh impossible. Based on these calculations you can confidently say the die is loaded and smack your friend in the head.

This is NOT influenced by the total amount of dice in the world, nor the amount of games played in the recorded history of mankind. Based on what you see right there you can assume something is amiss.

I'm sure you see the parallel with epics here!

I'm venturing a guess here: I think you both might be forgetting that getting epics (or throwing sixes) is completely independant of whatever else has happened before. You are right to say that the larger the sample-size, the more likely it is to see extremes, but that alone doesn't justify the extremes happening. Every player here is independant of all the others, so that's why above calculations and conclusions hold for a single player.

Something else came to mind;
There are mathematical grounds to (seriously) doubt the 25%, but that doens't mean there actually IS something wrong.

But simply put in above case:
The chance the die is loaded is 99,999999%, the chance the die isn't loaded is 0,000001%. Which is the 'safe' bet to make?

Last edited by Cregan; 10-24-2011 at 02:24 PM.
  #49  
Old 10-24-2011, 06:42 PM
Charista Charista is offline
Adventurer
 
Join Date: Jun 2010
Posts: 448
Default

No, you can't mathematically determine that he has a loaded die. All outcomes are possible, however unlikely. There is no point at which you can be certain the die is loaded.

The way to figure out if it is loaded is to take it away. Everyone uses that one die. Now it doesn't matter if it is loaded, because it's fair when you're all using the same one.

Your analogy does not hold to S&F item drops. We have no control over the die. We all use the same one. It cannot be loaded for or against one player.
  #50  
Old 10-24-2011, 10:26 PM
Bullbound Bullbound is offline
Supreme Adventurer
 
Join Date: Apr 2010
Posts: 3,208
Default

In addition to Charista's comment, at the same time you are assuming something that is mathematically possible is actually impossible. You have proven how unlikely it is, but that also proves how likely it is. You don't have an absolute certain. Also, folks have done polls to figure out the frequency of epics, and found it is indeed closer to 25% on average for many players and accounts.
__________________
[SIGPIC][/SIGPIC]
Please click "Thanks" if you found my post helpful
 

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -6. The time now is 03:04 PM.


Powered by vBulletin® Version 3.8.4
Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
Copyright by 'Playa Games GmbH'