TOPIC: The Problem with the Common Distribution...

First 80 bags...<br />one had 5 of a kind<br />one had 4 of a kind<br /><br />Second 60 bags<br />one had 4 of a kind<br /><br />at least 5 bags had 3 of a kind - I stopped counting at 5 bags...<br /><br />I'll let you know what is in the other batch if it makes a dif.<br /><br />is 5 of a kind likely to happen 1 bag in 200, or is my math as old as I am?<br /><br />G

<br />There was an experiment in one of my Statistics classes once where 1/2 the class was asked to write down a hundred "random" zeros and 1s.  01001101110001011010, etc.<br /><br />And the other 1/2 of the class was asked to generate a hundred random 0's and 1's using a computer.<br /><br />The Professor then went through and was able to easily pick out the real random from the selected random selections of 0's and 1's.  The main difference between the two sets of papers was the "rare" strings of many 0's or many 1's together.  In other words, the selected 0/1s usually never went longer than a string of 4 of either a 0 or a 1.  But in reality, randomness has a tendency to have a wider variation than we typically expect.<br /><br />I say this because I crunched the numbers for Random Common distribution and found that on average <br />36% of all packs will have a pair in it<br />5% will have two pair<br />2% will have a three of a kind<br />0.1% will have a four of a kind<br /><br />I'm not saying that their isn't a sorting problem but if you ordered 100 packs, then only 57 of those packs will have all unique commons and two of the packs on average will have 3 of a kind.   Further, receiving as few as 6 of one common and or as many as 32 is fairly typical.<br /><br />Statistics are funny sometimes.   =)<br /><br /><br /><br />

<br /><br />Thanks for the info, Douglas.  In my sample the non-unique packs was at 56.7% -- or about 13% higher than it should be.  That's enough to make sure it is done correctly.  I especially want to make sure the 3,000 player tokens we are giving out at purely random.

Jeff,<br /><br />Thanks for looking at the rares...I'll be interested to see what you find.<br /><br />Here are my facts:<br />60 bags - 6 +1 short swords.<br /><br />Maybe I am not recalling what I though I saw more than one person post...I though I heard folks say they got tons of one thing but little or 0 of others... Henwy, John...Mike?  <br /><br />I'm not complaining here - just trying to help improve IF improvement is found to be needed.<br /><br />Thanks again for looking - I know it is time consuming.

I would also assume that, if there was a sorting problem with the commons, the same problem would extend to the uncommons and rares.  Perhaps not, but it seems likely.<br /><br />I had 210 ten-packs.<br />I, too, had large stacks of some Reds and very small stacks of others.  I really don't know if that's correct statistically, but you can have my spreadsheet of what tokens I snagged in that 2100.  You can check it and decide for yourself if that appears correct.<br /><br />Thank you, Jeff, for listening to your customers and not taking our comments as idle chatter or whining.  I'm glad you looked into the "I got 3-5 of the same token in my ten-pack" issue.  It's good to know you look after your customers - and TD's reputation.<br /><br /><br />John<br />

<br />I think the COMMON distribution problem is much smaller an issue comparred to the RARE problem.  (At least I think I am hearing more people say things like - I got 7 "gauntlet" REDS but no "platemail"...)<br /><br />People are reporting things like getting 6-8 of an item in 60-80 bags...while at the same time they did not get a single Platemail or something...That seems against the odds or standard deviation thing.<br /><br />It is too bad that hand sorting has to be done to ensure better common distribution, but if TD will go to that much trouble for commons, perhaps we can address the RARES as well.<br /><br />It sure seems to me that IF a person buys 80-100 bags - (at least twice the number of rares in the set) they should have a complete set of REDS.  I've opened 140 bags and do not have a full set.  That seems more of a problem to me than the one bag I opened to find 5 grapple hooks or the two bags I opened to find 4 daggers and 4 short swords...<br /><br />Is there a way 80-100 bag lots can be created so that a person can be assured a full set of REDS?  I know this defeats the random factor, but you have to admit if 140 bags can't get a set - there is a problem leaving things up to random luck. (People will question how random it really was...)  Some if not most CCGs do this when you get an entire case - you usually get a full set evenly distributed and within a certain standard deviation.<br /><br />If fixing the Rares is not seen as a priority, I doubt I will understand fixing the commons...spare TD the time and headache of hand sorting for 2-3 bags with 3-5 commons.  If this happens 3-5 times in 150 bags - then the Coaches can have things on hand to give out to the person so afflicted...<br /><br />I think more focus should be on the Rares...<br />

<br /><br />Gary,  this is actually inline with random distributions.  I crunched the numbers and ran a simulation for 10 orders of 100 bags each.  <br /><br />On average, each of those orders were:<br /><br />missing 3 rares, had 1 copy of 8 rares, 2 copies of 12 rares, 3 copies of 9 rares, 4 copies of 6 rares, 5 copies of 2 rares, and 6 copies of 1 rare.  (Note: There is some rounding so it doesn't exactly add up to 100)<br /><br />But yeah, with the orders, 4 of the 10 people received 7 of one token and were missing at least one of the rares.  One unlucky soul was missing 6 Rares.  <br /><br />So based on comments, the Rare distribution matches what you would expect in a Random distribution.

Then in that case - I'm more for forgetting random distribution and adoption of CCG packaging...<br /><br />If you buy a "case" you should be assured of getting a set. (80 or 100 bags is a case in my opinion.)<br /><br />Not that trading isn't fun - but going CCG style would sure put an end to the lop-sided stacks...and deal with this little turnip:<br /><br />100 bags = a $1000 spend...what collectible game do you know of where people spend that much money and not get a full set of the rares?

<br />I had 210 ten-packs.<br />I, too, had large stacks of some Reds and very small stacks of others.  I really don't know if that's correct statistically, but you can have my spreadsheet of what tokens I snagged in that 2100.  You can check it and decide for yourself if that appears correct.<br />

<br /><br />John,<br /><br />Thats typical.  If you look at Random variations with a 210 pack order you might see something like this:<br /><br />Copies                0 1 2 3 4 5 6 7 8 9 10 11 12 13<br />number of times  0 1 3 6 4 8 6 6 3 2 1 0 0 0<br /><br />In other words, 1 copy of one of the Rares and 9 copies of 2 different rares.  Thats what you would expect.

<br />Then in that case - I'm more for forgetting random distribution and adoption of CCG packaging...<br />

<br /><br />CCG packaging is random too and you would see the same thing that you are seeing with True Dungeon rare distribution.  Believe me, I have tried to build sets before and it can be frustrating.  =|

I got 180 ten token packs.  I had been of the opinion that the distribution looked random to me, but given we are running though them, here the data from my rares:<br /><br />The image below is a histogram of how many times (Frequency) I got "Bin" numbers of the same token. (I.E. There is only one token I got nine of, so "Bin" 9 has a "frequency" of 1)<br /><br />The raw data was:<br /> <br /> <br /> <br /> <br /> <br /> <br /> <br /> <br /> <br /> <br /> <br /> <br />

Bin	Frequency
0	1
1	3
2	4
3	5
4	7
5	7
6	8
7	1
8	3
9	1

<br /><br /><br />[attachment deleted by admin]

I'm not sure all CCG packaging is totally random - I thought there were some games that would give you at least one complete set if you bought a box/case.<br /><br />Brian<br />

<br />Then in that case - I'm more for forgetting random distribution and adoption of CCG packaging...<br /><br />If you buy a "case" you should be assured of getting a set. (80 or 100 bags is a case in my opinion.)<br /><br />Not that trading isn't fun - but going CCG style would sure put an end to the lop-sided stacks...and deal with this little turnip:<br /><br />100 bags = a $1000 spend...what collectible game do you know of where people spend that much money and not get a full set of the rares?<br />

<br /><br />What you're describing is actually a very new innovation and not classic 'CCG style'. The vast majority of CCGs are still packaged in random distribution. Some small market ones have changed recently to guarentee no duplicate rares within a booster box, but that's only because they lack a strong player base and market. It's still rare to run into those games. I can only think of 3-4 off the top of my head.<br /><br />As for what game would you have to purchase 1000 worth and still not have a full set....clout comes immediately to mind. I considered getting into the thing but their distribution was insane. You would need to buy around 8 booster boxes just to have the _CHANCE_ to have a full set. That's assuming there would be absolutely no duping. With dups added in you would probably have to buy at least 30-40 to have even a moderate chance. Since their booster boxes were going at around $50 a pop, even wholesale, it's quite a bit more than 1k. <br /><br />As for the distribution this time around, I was missing 1 rare and didn't even hit the 1k mark in spending. That's just how the statistical cookie crumbles sometimes. As we discussed earlier with the Std Dev estimates, you would need to get huge swings to be in the realms of an outlier.

<br />Here are my facts:<br />60 bags - 6 +1 short swords.<br />

<br /><br />Just ran the 60-pack numbers and typically you would in fact get 5 or 6 copies of 1 rare, at least in 50% of the cases I ran.  :|